Embed Size (px)
Citation preview
Banking and Economic Growth: Case of the Gambia
Inaugural-Dissertation zur Erlangung des Grades eines
Doktors der Wirtschafts-und Sozialwissenschaften Der Wirtschafs-und Sozialwissenschaftlichen Fakultaet der
Christian-Albrechts-Universtitaet zu Kiel
Vorgelegt von M. Economics Bukhari. M. S. Sillah
Aus Wassu
Kiel 2005
II
Contents
List of Tables………………………………………………………………….……………..v List of Figures…………………………………………………………………………....…vi Abstract …………………………………………………….…………………………...…vii Dedication ……….. ………………………………………………………………………viii Acknowledgements …………………………………………………………………….…..ix 1 Introduction
1.1 Research problems ……………..…….………..………..……………..…….11
1.2 Objectives and contributions …………….…………………..……………….12
2 Background to the Gambian Economy
2.1 Introduction ……………………….………………….…...……………...…14
2.2 Currency Board ………………….………………….………………………17
2.3 Central Bank ...………………….……………….………………………......20
2.4 Commercial Banks ………………………….…………….……..………….32
2.5 Government Finance ….……………….…………..….……………..……...50
2.6 Trade and Industry …….………….…………………………………..…….55
2.7 Conclusions …………….…..………………………..………………..…….68
3 Survey of Theory and Evidence
3.1 Introduction ………………………………….….….……………………….71
3.2 Existential Theory ………………………….……..………………………...71
3.2.1 Information Economics ………………………………….…….………...……..72
3.2.2 Evolutionary Theory of Debt Intermediation ………..….………………..74
3.3 Empirical Evidence …………………..……..….……….….…………….....75
3.3.1 Neutrality of Finance ……………….…….………..……….…………………..76
3.3.2 Debt-Intermediation Retards Growth ……………………….………………….77
3.3.3 Finance Follows Growth……….….…………………………………………....78
3.3.4 Finance Causes Growth ………………….…..………….……………………..79
III
3.4 Conclusions ……………….………..……………………………………….80
4 Theoretical Framework
4.1 Introduction ……………..…………………………………….…….……...82
4.2 Schumpeterian Model ………………………………..……………………82
4.3 Broaddus’ Model of Efficient Allocation ………….…...…..…….…….....91
4.4 Keynesian Model With Credit Market ………………………..……….......94
4.4.1 Perfect Credit Market……………………….……….……….…….96
4.4.2 Imperfect Credit Market……………………………..……….…...105
4.4.3 Imperfect Credit Market, Money and
Credit are interlinked…………………………….……………116
4.5 The Behavioural Relationships…………………………………………...132
4.6 Tsuru’s Model of Financial Intermediation ……….…….……………….138
4.7 Conclusions …………………….…..……………….……………………148
5 Analysis of Survey and Empirical Results
5.1 Introduction ………………………………………………………............151
5.2 Survey Outcomes ……..….………………………………….…………...151
5.2.1 Bank Account Distribution ……………………………………..………………152
5.2.2 Reasons for Placing Money in the Banks ……………………..………………..153
5.2.3 Chance of Obtaining Bank Facility …………………..…………….…………..155
5.2.4 Reasons for Credit Application Being Approved………………………………156
5.2.5 Public Attitude on the Banking Services ……………………………………….158
5.3 Econometric Analysis ………………………………..……………............159
5.3.1 Unit Root Tests ………………………….………..…………………………….159
5.3.2 Banking and the National Output ………………………………………………162
5.3.3 Banking and Physical Capital ………………………………………………….168
5.3.4 Banking and Economic Efficiency …………………………………………….174
5.3.5 Banking and Economic Activities ……………………………...........................179
5.3.6 Broaddus’ Model ……………………………...………………………………..191
5.3.7 Banks’ Credit Market Analysis ………………..…………….............................193
IV
5.3.8 Schumpeter’s Credit …………………………………………............................206
5.3.9 Monetary Policy Competence …………….………..…………………………..208
6 Conclusions and Recommendations ………………………..………………….211
References ……………………….…………….…..….………………..…………...….216
Appendices …………………………...……………………..………………………….223
V
List of Tables Table 1: Unit Root tests. ……………………….……………………………….………...160
Table 2: Gross Domestic Product and Banking (unrestricted system)……..….…………162
Table 3: Gross Domestic Product and Banking (Restricted system)……………………..164
Table 4: Physical Capital and Banking (unrestricted system)……………………………169
Table 5: Physical Capital and Banking (restricted system)………………………………170
Table 6: Economic Efficiency and Banking (unrestricted system)………………………174
Table 7: Economic Efficiency and Banking (restricted system)…………………..……..175
Table 8: Economic Activities and Banking ( 7 variables, unrestricted system)………….181
Table 9: Economic Activities and Banking ( 7 variables, restricted system)…………….183
Table 10: Economic Activities and Banking ( 6 variables, unrestricted system)...………186
Table 11: Economic Activities and Banking ( 6 variables, restricted system)…………...187
Table 12: Broaddus’ Banking Efficiency (Lending and Deposit interest rates)………….191
Table 13: Supply and Demand Functions of Credit
(4 variables, unrestricted system)……………………………………………...195
Table 14: Supply and Demand Functions of Credit (4 variables, restricted system)……..196
Table 15: Supply and Demand Functions of Credit (5 variables, restricted system)……..199
Table 16: Real Money Demand Function (restricted system)…………..………………..202
Table 17: Transaction Cost Function (unrestricted system)…………..…..……………...204
Table 18: Transaction Cost Function (restricted system)………………………………...205
Table 19: Schumpeter’s Credit (unrestricted system)……………………………………206
Table 20: Schumpeter’s Credit (restricted system)……………………..………………..207
Table 21: Monetary Competence (unrestricted system)………………………………….209
Table 22: Monetary Competence (restricted system)…………………………………….210
VI
List of Figures Figure 2.1: Funds of the Central Bank…………………………………………………….21
Figure 2.2: Assets of the Central Bank………………………………………………........23
Figure 2.3: Gambia Inflation versus UK Inflation………………………………………...27
Figure 2.4: Exchange rate Index, Gambia Inflation Rate
and the Money Growth rate...............................................................................28
Figure 2.5: Foreign Assets versus Domestic Credit……………………………………….30
Figure 2.6: Sources of Commercial Banks’ Funds……………………………………......36
Figure 2.7: Loss or Gain from the Commercial banks’ Balance Sheets………………......41
Figure 2.8: Lending Interest Rate, Deposit Interest rate and the Spread………..………...43
Figure 2.9: Distribution of Commercial Banks’ Assets…………………………………...45
Figure 2.10: Commercial Banks’ claims on the Economic Sectors……………………….47
Figure 2.11: Customs’ Contribution to Government Finance……………………………..52
Figure 2.12: Budget Deficit/ Surplus……………………………………………………...54
Figure 2.13: Groundnut Share in the Export………………………………………………56
Figure 2.14: Current Account Activity…………………………………………………....63
Figure 2.15: Log GDP over a Century…………………………………………………….65
Figure 5.1: Bank Account Holdings by the Public…………………………..………......153
Figure 5.2: Reasons for Placing Money in the Banks……………………………………154
Figure 5.3: Chance of obtaining Bank Credit………………………………...………….156
Figure 5.4: Reasons for having qualified for Bank Credit……………………………….157
Figure 5.5: Ratings of the Commercial Banks…………………………………………...158
Figure 5.6: Impulse Responses of LRY in VAR system………………………...………167
Figure 5.7: VEC Impulse Responses of LRY…………………………………………....168
Figure 5.8: Impulse Responses of LRPK in VAR system……………………………….172
Figure 5.9: VEC Impulse Responses of LRPK……………………………………..……173
Figure 5.10: Impulse Responses of REFL in VAR system………………………………177
Figure 5.11: VEC Impulse Responses of REFL…………………………………………178
Figure 5.12: VEC Impulse Responses of Economic
Variables versus Banking Variables ……………………………………...190
VII
Abstract
This research advances four theoretical approaches in an attempt to relate the banking activities to the real economic activities. It starts with the Schumpeter’s circular flow of creditary production that argues that the banks start the production cycle for offering the credits that enable the entrepreneurs to purchase labor and capital. The banks increase the savings mobilization and allocate the scarce savings to the productive investments. Second, we develop a benchmark from the Broaddus’ competitive model to analyze the competitiveness and efficiency in the banking industry. The third theoretical approach incorporates a credit market in IS/LM analysis and discusses the credit constraint as an analogy to a quantity constraint of the neo-Keynesian theory. This approach also analyzes the impacts of the interaction between the monetary policy and the fiscal policy on the endogenous macroeconomic variables. We finally modify Tsurus’ model to show that banks’ mobilized savings could be spent on the maintenance of the banks rather than being channeled to the productive investments, and as a result a bank-based economy could perform worse than a nonbank-based economy. We then estimate and test the hypothetical relationships between the banking and the real economic activities, and estimate and analyze the banks’ credit market functions and the credit constraint hypothesis. We use Johansen Vector Error Correction Methods, VECM, for all the estimations, and hence the analysis is focused on the long run relationships and the adjustments towards the equilibrium. We also conduct an explorative survey into the public’s relationships with the banks. The research finds that the banks’ credit to the private sector is vital for the real economic activities of output and capital accumulation, it is found to be Granger causal for these activities, and it is a weakly exogenous variable in the equilibrium systems that do not include private sector investments; while, the bank liabilities are found to be an endogenous variable. The interest rate is found to slightly influence the decision of the public to save in the banks; the public see the credit facilities biased towards the consumption financing than investment financing. The lending interest rate has a small effect on the credit supply. The banks’ credit supply is inelastic with respect to the lending interest rate, and there is weak credit constraint in the credit market. The banking industry is found to be uncompetitive and inefficient; and the increased transaction costs of the bank credit market are associated with increased prices in the economy. These increased transaction costs are also found to cause the public to hold increased real money balances. The banks’ credit and lending capacity are found to depend on the private sector investments, increases in the private sector feed back onto both the banks’ credit and lending capacity. The banks lag behind the developments in the private sector; thus they are not promoters or engines of growth for the private sector.
VIII
Dedicated to my father, may his soul rest in peace.
IX
Acknowledgements
I wish to offer here my sincere gratitude to people, who stood by my side when things were
odd. They rendered invaluable assistance to me that I could stand firmly on my feet
throughout this program. I am very much grateful to Professor Dr. H.-W. Wohltmann,
Professor Dr. Till Requate and Frau Scholz for the kindness they showered on me during
my early days into the program. I thank them very much. I have no words to describe the
sincere and painstaking supervision I received from Professor Dr. Wohltmann. He has
accommodated me amply and cordially, time has never been a problem, he would welcome
and attend to me any time I came to his office. His suggestions and comments have
fundamentally molded and made this research possible. He has been greatly considerate of
my time limitations, and thus he was always on time to read and comment on my drafts. I
am specially indebted to him. I am also indebted to Professor Dr. Helmut Herwartz for
supervising the econometric analyses. I acknowledge, with thanks, his kind and
constructive supervision. Grateful acknowledgement is made to the special comments and
suggestions offered during the presentation of the research at the Institute of Statistics and
econometrics, CAU, Kiel. I also acknowledge the technical support the PhD program,
Instituet fuer Volkwirtschaftlehre and the Library of Kiel Institute of World Economics
have adequately made available to me. Special thanks are made for the kind supports I
received from Evangelische Studenten Gemeinde of Kiel University and Africanischer
Muslimischer Verein Kiel. My great gratitude goes to DAAD. The Award I received from
DAAD was a turning point in my life. Without the Award I could not have been able to
continue in this PhD program. I thank them very much for all the support. There were
friends, colleagues and family members, who kept motivating me to keep up the
momentum. I thank them all and acknowledge their help.
- 10 -
1 Introduction
Paul Romer (1993) says “A nation that lacks physical objects like factories and roads
suffers from object gap. A nation that lacks the knowledge used to create value in a modern
economy suffers from idea gap.” To complete, a nation that lacks finance to erect physical
objects and produce knowledge suffers from all gaps. A well-functioning banking sector is
sufficient to create economic modernization. Alexandra Hamilton (1781)1 argued that
“banks were the happiest engines that ever were invented for spurring economic growth”.
That is, there is a causation that runs from banking sector development to economic growth.
The recent experience in Eastern Europe and Asia has shown that countries that moved
quickly to fix their banking industry were able to achieve a sustainable rate of growth and
new job opportunities than those that did not. But why this hypothesis did not hold for the
case of the Gambia, with some banks existing in the country for over a hundred year. Are
these banks really financing the domestic economic growth? Or what went wrong with the
banks including the central bank in the country? These are some of the questions this
research intends to tackle. The research will span from the public perspectives about
banking to banking activities and the impact of those activities on the economic growth.
The research advances four models that attempt to link the banking activities to the real
economic activities. It starts with Schumpeterian circular flow of creditary production that
argues that banks enable the entrepreneurs to carry out new economic activities, and that
any new development that has no previous products must be incubated by a bank credit.
The second model gives an economic representation that investigates the efficiency and
competitiveness of the banking industry. The third model incorporates a bank credit market
in an IS/LM analysis and discusses the credit constraint, and the fourth model investigates
the banking role in the savings mobilization and allocation. The research attempts to
analyze three problems, which are defined in the next section.
1 A quote from Levine, Ross; Loayza, Norman; and Beck, Thorsten (2000), Journal of Monetary Economics 46, pp. 31 – 77.
- 11 -
1.1 Research Problems
1.1.1 To inquire historically in the Gambian economy with emphasis on the banking
industry and financial system. This is to attempt to identify some major
parameters of the economy and how those parameters have been changing over
time.
1.1.2 To investigate the public view about the banking industry in the country and
attempt to formulate the reasons for which the public hold bank accounts and
the reasons for which the public get the bank facilities, and to improve the
understanding about the types of projects the banks tend to finance. This is done
by conducting an explorative survey, the survey questionnaire sample is in
Appendix C, and the survey outcomes are discussed in section 5.2.
1.1.3 Banking and the economy: I attempt to formulate models in which the role of
the banks can be identified in the economy and the ways in which the banks can
contribute to the economic growth. Hypothetical models are developed in which
the banks can start the circular flow of the economic growth and development,
and how the efficient performance of the banks can be channelled to promote
the economic activities. I develop a bank credit market that can interact with the
other markets to determine the national output, transaction costs, interest rate,
prices and the credit quantity of the banks. I also attempt to formulate models in
which the bank credit can affect the private sector investments and help
determine some other endogenous variables. I then analyze statistically and
econometrically the hypothetical models over the period 1964 to 2002 in order
to spot the linkages between banking and economic activities.
- 12 -
1.2 Objectives and Contribution
The Gambia has no abundant natural resources; but this is no excuse for underdevelopment.
In fact, it has been observed that countries with abundant natural resources have tended to
grow less rapidly than natural scarce economies. Sachs and Warner (1997)2 found a
negative relationship between natural resource abundance and economic growth,
confirming the old notion of Dutch Disease that natural resource booms crowd out other
sectors that have positive externalities in the form of learning by doing (Boschini,
Pettersson and Roine, 2003). Recent findings by Welsch (2004) and Esanov el. al(2004)
have also point to the validity of resource curse. The other views suggest that whether the
resource is a curse or not depends on the interaction between the institutional setting and
the type of natural resources (Boschini, Pettersson and Roine, 2003, and Hausmann and
Rigobon, 2002). Hausmann and Rigobon found that financial market imperfection worsens
resource curse. This is the point we like to develop in this research; we are not here to write
on the natural resource-growth relation, but finance-growth interaction. It is finance, bank
credit, that matters to the economic activity; if there is a finance gap, then there will be gaps
everywhere in the economy. With sound banking system, we should be sure of economic
modernization. The resolutions to the research problems will contribute towards the long
debates and on-going researches that the Gambia should be a financial centre and a trade
gateway of Africa, particularly West Africa. It is argued that because the country is small
and has no natural resource, it should specialize in banking, finance and inter-port trade.
The debates have been going on without any scientific research on how current banking
activities are affecting the economy. I believe, before any strategic policy to make a country
a financial centre, the current banking sector should be thoroughly appraised against the
long term economic goals of the country.
This research will be first of such kind that will explore what is going inside the banking
sector, how that activities affect the economic growth and modernization and what could be
recommended for policymakers to make the country a financial centre. The research is
organized as follows, Chapter 2 reviews the Gambian economy with an emphasis on the
financial sector; Chapter 3 presents a brief survey of the theory and evidence for the role of
2 A quote from Manzano and Rigobon, Resource Curse or Debt Overhang?, Working Paper 8390, National Bureau of Economic Research, 1050 Massachusetts Avenue, Cambridge, MA 02138, July 2001.
- 13 -
the banks in the economy. Chapter 4 discusses the theoretical framework, where four
approaches are presented to analyze the banking contributions to the economic growth and
development. It also defines the behavioural relationships to be estimated and the variables
in each relationship. Chapter 5 presents and analyzes the empirical findings, and chapter 6
concludes the research and derives the recommendations. There are three appendices after
the list of references, Appendix A gives the results of order selection, co-integration rank
tests and some Granger causality tests. Appendix B tabulates the research data and
Appendix C gives a sample of the survey questionnaire. The lists of figures and tables are
given in the table of contents.
.
- 14 -
2 Background to the Gambian Economy
2.1 Introduction
Pre-colonial Gambia witnessed a variety of currencies for trade purposes. Gold,
Cowries, Strips of Cloth, Copper and Iron rods, were widely circulated for facilitating
liquidity and effecting exchanges of goods and services through the free market forces.
The simultaneous existence of these different currencies produced a complex currency
system in the country. However, each currency served some specific purposes. For
example, Copper rods were used for controlling liquidity because they were available in
small denominations. Similarly, cloth money or currency for its widespread availability,
easy valuations, was used for small trade transactions up to the colonial times. Gold and
Cowries, for their durability and high valuations, were used for savings, high valued
transactions, rituals and jewelleries. With the advent of imperial missions of France and
Britain, the monetization and the integration of the local economies into the capitalist
world were introduced. The introduction of the British currency was on the basis that
local currencies were not convertible into international currencies, and that some
currencies like cloth money had prohibitive transportation costs, and that all the local
currencies encouraged barter trade. Nevertheless, due to the miscalculations and the
pursuit of self interests by the colonial powers, the introduction caused some economic
miseries for the local population. The local currencies continued to exist along side the
colonial monies of British silver coins and French Five Francs until 1891. In as early as
1891 with the prohibition of the use of all local currencies, the colonial monies became
legal tenders. The colonial power then created and used a banking system to help
supply and distribute the British currency. Thus, the British Bank of West Africa (later
know as Bank of West Africa) was established in the same year 1891 to serve as both
central and commercial bank for the British colonies in the West Africa. The Bathurst
(now Banjul) branch was opened on September 8th 1902. “All the banking businesses
are transacted by this bank without charge, except in certain specified cases such as
remittances, overdrafts if desired, and a defining of the current account below ₤400, in
the last two cases the interest permissible being at the lowest rate the bank would
- 15 -
charge its most favourable customers”3. This was an agreement entered into by the bank
and the Government of the Gambia in 1902; it was aimed at monetizing the economy
and making available cheap financial services. Hut and yard taxes introduced by the
colonial government also helped institute the colonial monies in the Gambia. The
introduction of British currency and the prohibition of local currencies had severe
impact on the local economy and retarded the growth. The local merchants were
rendered poor because they could not redeem their local currencies at the British Bank
of West Africa; similarly, all ordinary people who were having cloth pieces, iron or
copper rods just had to consume them, or kept them until they got spoilt or lost, but
never to be redeemed for British currency. Worse still, the local merchants and
population had to pay all series of taxes in the new currency which was not widely
available in their hands. So, they were forced to be farmers not of their own any more
but of the colonial governments who wanted them to grow groundnuts, which the
colonial monopolists offered to buy. The economic balance was then changed; people
were forced to migrate and settle around the British administrative head quarters and
branches to get colonial jobs and got paid in the new currency. The breed of merchants
and administrators of Indians, Lebanese and Hakus emerged. These people were
brought by the colonial masters because they had the European taste and education and
could be trusted to run the protectorate. The Lebanese and Indians engaged in import
trade; they promoted the European goods and taste on the local markets; and the
colonial masters accorded them favours and opportunities, because they were seen as
opening new markets for European products and taste. Thus, no efforts were made to
manufacture export products or build import substitution products. This further
impoverished the local people, who had to rely almost entirely on the import goods
even for some basic subsistence needs. While, the demand for local products continued
to fall in the face of the European goods and with the fact that those who had the
colonial currency were in the metropolis and around the administrative areas and these
people were already undergoing cultural transformation and assimilation into their
masters’ culture, so, they would prefer to buy the European taste. The other negative
3 The Gambia Colony and Protectorate: An Official Handbook, Francis Bisset Archer, Frank Cass & Co. Ltd 1967, p. 272
- 16 -
impact brought about by the introduction of the British currency was that, during the
times of the local currencies, local merchants and ordinary people were their own
banks, they hoard the number of cloth pieces, gold, iron and copper rods they got; they
subsequently applied the same mentality on the new currency, by hoarding as much as
they could, kept under pillows, praying mats and some people resorted to burying them
for safekeeping. As a result more currency was kept as cash holdings in the individual
hands instead of as deposits to be channelled back to economy as investments. Many
people ended up as debtors to the money lenders who were charging 100 percent
interest rate. This drove out local investors from the market. Even the Gambia Co-
Operative and Central Banking and Marketing Union Ltd were having a daunting
challenge to conduct operations due to lack of local savings. This was highlighted in the
Second Report of International Labour Office, Geneva 1965, on the Co-operative
Banking in the Gambia:
“ … The Apex Bank depends largely on the provision of short term credits by the Bank of
West Africa, B.W.A, to carry out its operations. The overdrafts with the Bank of West Africa are
guaranteed by the Government. The expansion of the Apex bank’s activities leads to a growing
dependence on such government-guaranteed overdrafts from the B.W.A., should B.W.A. for some reason
be unable to grant further credits to Apex Bank, the latter would almost immediately be unable to
continue its operations. It is again recommended that the co-operative movement should make every
effort to build up capital by attracting savings and other deposits in order to avoid borrowing from
B.W.A.”
Unfortunately, the recommendation was not heeded of quickly; more currency continued to
be held in individual hands, spent on European taste, and ultimately withdrawn from the
economy. Apex bank continued to borrow at increasing interest rates, which could not last
indefinitely; then, it resorted to buying the produce on credit forcing the farmers to borrow
from money lenders to pay taxes, meet subsistence needs and prepare for the next farm
season. This pyramid could not later be sustained, but it had to collapse with the farmers
buried under its debris. This caused further migration of the rural population to the
metropolis, which was seen as a solution for obtaining the colonial currency; with those left
behind in the up-country totally dependent on those lucky ones who could make it to the
metropolis. This created economic insecurity and instability in the entire country.
- 17 -
2.2 Currency Board Episodes
The arrival of British introduced the British coins such as copper coins and silver coins,
which were exclusively imported by British Bank of West Africa. By 1900 the
monetization of West African economies was on a high gear and the demand for British
silver coins in West Africa exceeded that within the British economy itself. By 1912,
according to Schuler (1992), “the seigniorage of British silver coins was 165 percent of the
value of their silver content, and this situation agitated the West African Colonies and they
demanded the British Treasury to share the seigniorage with them or allow them to
introduce a separate West African currency”. Their demand was no heeded of until 1913
when the West African Currency Board was established and allowed to issue its silver coins
and notes. Currency board, unlike a central bank, holds at least 100% foreign reserves
against all the notes and coins it issues. . The currency board has no monetary policies; its
job is to respond to the supply and demand forces of the foreign exchange market. The
West African Currency Board was the first typical currency board system. It enjoyed
impressive stability and generally good macroeconomic performance. It started to crumble
down as the colonies gained their independence. Ghana was the first to throw it away as it
got its independence in 1957 and established a central bank in 1958 by converting a
government commercial bank to a central bank. Then Sierra Leone and Nigeria followed
Ghana, as they too got their independences. Meanwhile, Gambia maintained it, and it
became The Gambia Currency Board even after the independence until 1971. It will
however be difficult to judge the performance of currency boards over the period 1913 to
1971 as an effort to compare that with the performance of the central bank afterwards. The
era of currency board is a complex one due to the colonial administration that linked the
domestic economy completely to the British economy. It though connected the domestic
monetary system to that of colonial masters; there is a merit to shed some light on its
episodes. The first episode ran from 1913 to 1963; that is the period when the board is a
British West African Currency Board. The second period ran from 1964 to 1971; that is
when it became a Gambian Currency Board. The data on the first episode is difficult to
extract or evaluate its marginal effects on the health of the Gambian economy.
In the first episode, the reserve ratio and assets was 110% of British Pound Sterling assets,
it reduced to 100% in the second episode. The exchange rates in both periods were one
- 18 -
West African Pound or Gambian Pound to one British Pound plus 0.5 %. Immediately after
the currency systems, the Gambian currency depreciated 67% against the Sterling Pound;
and until today, any currency issued by the central bank is worth less than the Sterling.
Contrary to other currencies of former British colonies, such as Hong Kong, Brunei,
Bermuda, that still maintained currency boards,” no currency still issued by them is worth
fewer Sterling today than in 1950”4. The domestic economy enjoyed low levels of inflation
during the whole episodes of the currency board; the inflation averaged 1.5% per annum;
while the GDP grew on average 4.5% per annum. The aftermath of currency board has
witnessed increasing levels of inflations coupled with declining or constant real per capita
growth rates. From 1972 to 1989, the inflation averaged 15% per annum, and the GDP
averaged 1.1% per annum5.
The traditional functions of the British West African Currency Board were purchases and
sales of sterling, and the management of overseas investments. The West African pound
was issued as the Board purchased sterling and withdrawn as the Board sole sterling. This
activity was purely determined by the demand and supply forces of the market. The volume
of trading determined the volume of West African currency in the circulation. The Bank of
British West Africa acted as an agent in issuing and redeeming the currency. The other
traditional function was the management of its overseas investments, which were mostly
held in UK assets. The money demands in the member countries (Nigeria, Southern
Cameroon, Ghana, Sierra Leone and the Gambia) were not of course perfectly positively
correlated; thus, at any given time, the board had sufficient redeemed currency to be
invested in high-yielding UK assets that could earn it income to cover operating costs of the
Board and credit the profits to the Government accounts of the respective members.
Nevertheless, a high portion of the overseas investments were in the UK government
treasury bills, which were highly liquid and could be easily liquidated to meet the seasonal
money demands arising from the harvest and marketing of crops in the agric-economies of
the member countries.
When the Gambia Currency Board was established in 1964, after the other members
abandoned the board, it found itself as a financial arm for the one-crop economy of the
4 Schuler, Kurt A., (1992), Currency Boards 5 Ibid
- 19 -
Gambia. The ordinary revenues of the Government could hardly meet the recurrent
expenditures, and that grants and loans were always required to build new capital work or
expand the amenities and services. The Co-operative Central Banking Union, the sole
financier of the business of Gambia Oilseeds Marketing Board, which was the sole
purchaser of the economy’s sole industry’s output, groundnut, had to borrow from the Bank
of British West Africa at high interest rates or via Government guarantees. With collapsing
sole industry and strained Government budget, the currency board was mistaken for the
currency printing machine, and the acts of the Currency were quickly amended to
incorporate the following:
1. Discounting and rediscounting of inland bills of exchange and promissory notes
payable in lawful money of the Gambia arising out of the marketing of agricultural
produce.
2. Discounting and rediscounting treasury bills of the Gambia Government payable in
Gambian pounds.
3. Granting advances against the security of promissory notes issued by the
Government and payable on demand6.
The Board could also undertake fiduciary lending in respect to financial assistance in the
marketing of the staple crop, groundnut, and the extension of credit to the Government.
The amendments in practice disabled the Gambian Currency Board and transformed it into
more than a central bank. Even countries then that opted for central banks did not have such
functions like direct lending to the Government. In addition, to purchases of sterling, the
Board now had to issue currency against the acquisitions of local credit instruments issued
by the government, commercial banks, local merchants and Co-operative Central Banking
Union. Consequently, the sight liabilities of the Board increased tremendously forcing it to
reduce its external cover from 110% to 100%; and within two years of its existence, the
minimum external cover was just around 70%.
6 Annual Reports of The Gambia Currency Board, 1964 – 1968, Banjul.
- 20 -
2.3 Central Banking System
It was argued, as Schuler (1992) has narrated, that currency boards did not allow the
policymakers a room to influence the national employment and undertake development
projects and did not satisfy the Government need for money; that currency boards did
transmit external shocks directly into the domestic economy due to the nature of its foreign
reserves and the exchange rate regime it had to adopt, which is a fixed or managed system.
Thus, the Central Bank was established to pursue price stability, high employment and
growth. It would also act as an economic adviser to the government and function as a
national bank that would conduct monetary policies to help channel government
development and stabilization objectives into the economy.
In this section, we analyze the balance sheets of the Central Bank of the Gambia in the light
of the above stated objectives, and evaluate its performance over the period from 1971 to
2002. The funds of the central bank consists of reserves, Government deposits, foreign
liabilities, capital and other items, which include the valuation adjustments. The conditions
of the funds was stable, each of these components closely maintained its share in the total
funds from 1964 to 1981/1982.
- 21 -
Fund
s o
f the
Cen
tral
Ban
k
-250
-200
-150
-100-5
0050100
150
200 19
64
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
percentage of the total liabilities
Res
erve
Gov
t.Dep
For
eign
Lia
bilit
ies
Cap
ital
loss
or
gain
- 22 -
The Government deposits slightly declined from 28.9 per cent in the year the central bank
was inaugurated to 3.4 per cent in 1980, it later rose somewhat. The central bank did not
hold foreign liabilities until 1976. The reserve was also stable before 1981/1982; in that
period, it accounted for 50 per cent of the total funds of the central bank. The stability of
the funds was disturbed by two major events that rocked the Gambian economy with a lag
effects spanning from 1981 to 1992. In 1981, an attempted coup deta foiled by Senegalese
intervention triggered the ready-to-explode problem of the central bank. The central bank
replaced the Gambia Currency Board at the time the asset holding of the board has dropped
from the required 100% to an unsustainable level of 67% of the total assets; and the rest of
the assets were claims on the Government and the deposit money banks. Thus, the central
bank did not come into being because of some genuine economic reasons as was claimed; it
came because the Currency Board has lost its meaning and was crippled by asking it to fine
tune the economy by giving credits to the Government, the deposit money banks and the
groundnut merchants. This behavior was completely contrary to the meaning of the
Currency Board; even the then central banks were hardly found involved in such cases. The
problem became amplified with the advent of the central bank. The bank pegged the
Gambian currency to the British pound, while the fiscal authorities introduced the measures
of price controls. These measures were taken to combat the 1970’s inflation blamed on the
energy crisis. But that inflation could be also blamed in part on the unstable money supply
of the country. The money supply was highly variable, while the government budget deficit
was growing widely. Thus, pegged currency rate could not last long in the face of high
variable money supply and increasing budget deficit. This was combined with repressed
and low deposit rates and fixed lending rates. These pressures have created implicit
inflation rates higher than that of the United Kingdom, to whose currency the Dalasi was
pegged; thus the Dalasi had to depreciate against the sterling pound. Repressed interest
rates, managed exchange rate, increasing budget deficits and variable money supply created
high uncertainty in the economy, it did not explode because the country risk was almost
zero. The 1981 attempted coup ignited the explosion as the country risk was added to the
already unbearable economic and financial uncertainty. In 1985/1986, the peg was
abandoned, and the banking market was somewhat liberalized, and the fiscal authorities
- 23 -
Figure 2.2
Ass
ets
of th
e C
entr
al B
ank
01020304050607080
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
percentage of the total assets
Cla
ims
on p
rivat
e se
ctC
laim
on
DM
B
- 24 -
also abandoned the price controls. The central bank changed its policy from boosting
lending base of banking sector via credits to reducing the paper money in the economy via
reserve increases and sale of Government and central bank discount bills. The chart above
illustrates. The Government also reduced its role in the economy; the Government
enterprises such as port authorities, national water and electricity company and the
telecommunications company were let away to run semi-privately, known as parastatals. It
was painful adjustment; the Government laid off some civil servants, and attempts were
made to balance the budget, and the intermittent surplus balances were deposited in the
central bank to finance its fund base that has been wiped out by foreign exchange losses
and valuations. The valuation losses stood at 224 per cent of the total liabilities of the
central bank in 1987. The gain from the adjustment was immediately reflected in the
banking sector. It has become able to raise funds higher than before from both local and
foreign sources. The increase of the nominal interest rates attracted the inflow of the capital
that helped offset greatly the current account deficit. The private sector was activated as the
government reduced its role in the economy; the trade and re-export trade flourished.
However, the impact of the gains on the general welfare was not felt. The parastatals
focused on growth not development; so, they grew pyramids. The private sector in response
to the shortage of goods created by the previous policies of price controls, State department
stores and exchange rate pegs, they focused on quick refilling via imports. Thus, this sector
built its nests along the administrative and merchandise ports. It paid no attention to the
majority of the people, who depend on the groundnut industry. The purchasing power of
this people kept dwindling at every trade season as synthesized import goods displace their
produce. Some decide to migrate to the urban areas but only to add to already unemployed
secondary and high school graduates there.
Where was the central bank in the midst of these economic tumults and imbalances? The
central bank was busy going back to the era of its predecessor, the Currency Board. The
central bank lost almost all its foreign assets during the crisis; the foreign assets stood at
around 2.5 per cent of the total assets from 1983 to 1985.
- 25 -
Figure 2.2 (continued)
asse
ts o
f the
Cen
tral B
ank
020406080100
120 1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
percentage of the total assets
Fore
ign
asse
tsC
laim
s on
Gov
tC
alim
s on
Offi
cial
ent
ities
- 26 -
But suddenly after 1985, it increased foreign asset holdings to 46 per cent in 1987, and 87
per cent in 1991 which is higher than that of the Currency Board in 1968. Throughout
1990’s and until 2002, the foreign asset holdings remain very high accounting on average
for 86 per cent of total assets of the central bank. It cannot be a currency board and central
bank at the same time; so, what does the Central Bank of the Gambia pursue? Does it
pursue price stability, or exchange rate stability, or maximum employment, or banking
supervision and stability, or conflict avoidance with the fiscal authorities. Pursuing multiple
goals causes a central bank to lose its independence and to more exacerbate the economic
shock than to achieve the goals. The central bank of the Gambia tends to pursue multiple
goals and to accommodate the Government needs. This has rendered it ineffective in
achieving many of its goals. Prices have never been stable; they are either controlled as was
the situation in 1970’s to mid 1980’s, or they are left to spiral up very rapidly as has been
witnessed from 1990’s to date. The inflation is almost uncontrollable, but the measure of
inflation does not seem to say so. According to the Gambia CPI, the domestic inflation
from 1994 to date is either equal to or less than the UK inflation. This theoretically implies
that the Gambia currency must stay unchanged or appreciate against the UK currency. But
that was not the case; in fact, the Gambia currency depreciated more than 50% against the
UK currency. Thus, this can cause a researcher to suspect that the Gambia CPI does not
correctly and technically capture the general price changes in this period. In late 1960’s and
early 1970’s when the UK inflation was higher than Gambia inflation the Gambia currency
was revalued against the pound sterling from GD5/1₤ to GD4/1₤. Furthermore, it is a
forgone conclusion in the economic literature that inflation is a monetary phenomenon in
the long run. This does not seem to hold in the Gambia. The contemporaneous correlation
between money supply growth and the inflation is 0.033, one lag is 0.372, second lag is
0.167, third lag is 0.141 and fourth lag is 0.092. The money supply growth is highly
variable and has an increasing trend and the exchange rate index has been depreciating;
while the inflation rate measured by the CPI looks steady and somewhat falling. The central
bank might have been hitting a wrong target of inflation. Figure 2.3 graphs the Gambia
inflation and the UK inflation, and figure 2.4 presents the money supply growth rate, the
inflation and the nominal exchange rate index of the Gambia.
- 27 -
Figure 2.3
-100102030405060
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
UK
Infla
tion
Gam
bia
Infla
tion
- 28 -
Figure 2.4
-60
-40
-20020406080
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
EXH
Rat
e In
dex
Gam
bia
Infla
tion
Mon
ey G
row
th ra
te
- 29 -
It is also the responsibility of the central bank to supervise the banking sector and maintain
its stability. The central bank conducts open market operations using its own discount bills
and the Government treasury bills with largely the banking sector. The bank’s own fund
can be used for monetary management but the Government funds are directed by the fiscal
authorities, and the objectives of the latter funds normally overrun that of the former, and
when conflict is imminent, the central bank pursues the goal of conflict avoidance. The
bank acts also as a lender of last resort. Similarly, this operation can conflict with price
stability objective. Releasing funds to insolvent banks will derail the bank from its money
supply target. If it pursues maximum employment objective together with price stability
and other objectives mentioned in the beginning, a conflict will arise. In a supply shock
situation, for example, which is common in the domestic economy, that causes the actual
output to fall below the potential output, any action by the central bank to stimulate the
aggregate demand will result in inflation; thus, the price stability target will be missed.
Likewise, accommodating the Government needs either by letting the fiscal authorities to
use the sterilized funds or maintaining repressed interest rates for the Government
borrowing will cause the central bank to miss its other targets such as price stability and
exchange rate stability. Looking at figure 2.5 below, we get a glimpse into what objectives
the central bank has been pursuing. From 1980 to 1990, foreign funds started flowing into
the domestic economy, and the central bank responded by sterilizing the inflows in the
form of increased domestic credit; this resulted in a negative relationship between the
foreign reserve and the domestic credit. This, according to Roubini ( 1988 ) happens when
“a central bank cares more about interest rate smoothing objective relative to foreign
exchange reserve stabilization”. This period witnessed some lifting of financial restrictions
on interest rates and banking operations in general; the soaring interest rates after the lift
attracted foreign funds, but the central bank was cautious about high interest rates lest they
harm the economy; and thus it was more concerned about interest rate smoothing relative to
other objectives.
- 30 -
Figure 2.5
-100
0
-5000
500
1000
1500
2000
2500
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002million Dalasis
Fore
ign
Asse
ts (N
et)
Dom
estic
Cre
dit
Res
erve
Mon
eyC
urre
ncy
outs
ide
DM
B
- 31 -
The pattern has changed from 1990; both domestic credit and the foreign assets have been
increasing, implying a positive relationship, which means the central bank has become
more concerned with foreign exchange reserve stabilization relative to interest rates
(Roubini, 1988). Both the central bank and the commercial banks have been accumulating
foreign assets. Smoothed or flat interest rates coupled with falling export funds of the
groundnut industry have made the domestic economy unattractive to the foreign funds;
while increasing Government debts and debt services and increased demand for imported
food stuffs, building materials and high rates of migration from the country have put
pressure on both the central bank and the commercial banks to hold increasing foreign
assets in order to meet the demands for foreign exchanges stemming from imports, debt
services and migration. This combined selling of the domestic currency resulted in its
depreciation. The central bank responded by backing the currency in the form of increased
interest rates, and increased domestic credit, and recently by asking the banks to increase
their minimum capital requirements; the long term objectives of price stability and full
employment have already gone off track. The depreciated currency from its 1990’s level
combined with high interest rates have already been translated into high prices and low
private investments signaling a long term difficulty, especially if the earnings of the
accumulated foreign assets fall short of meeting the high domestic interest obligations. That
is, if the increased domestic credits are not in profitable and tradable economic activities
that can more pay the interest obligations, then theoretically the domestic economy must be
asked to liquidate, the currency will further depreciate, prices will go up and employment
will fall, the economic authorities must examine carefully the situation of the country’s
debts and cut down the interest rates. The current discount rates may even satisfy the
banking industry in meeting their deposit and operational obligations; thus rationally telling
the bankers to just put their funds in the treasury bills and central bank bills and then wait
for the maturity. This will deny funds to the productive sectors of the economy; the
discount rates must not be higher than the expected returns of the least risky economic
activity in the country.
In chapter six, we will see empirically how the Central Bank monetary policy has
performed over the study period.
- 32 -
2.4 Commercial Banks
Until 1973, banking and financial institutions were rudimentary, and the general public
knew nothing about them. From 1973, when the Financial Institutions Act was passed into
law, to 1986, the banking performance was precarious, it is only very recently that banks
have been able to somehow position themselves in the market, but still their impact on the
welfare of the general public is questionable.
The first bank established in the Gambia was the Government Savings Bank on 1st January
1886. It operated under the treasury department of the Government, and accepted limited
deposits from the public. It did not engage in business transactions; so, it had a limited role
in executing financial facilities for the entrepreneurs. In fact, its liabilities were under the
Government balance sheets. Unfortunately, these liabilities were in many years until 1964
invested in foreign assets, mainly British assets. The incomes generated from these assets
were used to pay the deposit interests, and the balances were credited to the Government
accounts.
Later on , when the Central Bank of the Gambia was inaugurated, the importance of the
Gambia Savings Bank declined, because the central bank could act as a better bank for the
treasury department than the savings bank; thus the Gambia Savings Bank was phased out.
The Bank of British West Africa ( Bank of West Africa changed to Standard and Chartered
Bank, and then today is known as Standard Chartered Bank) opened its branch in the
Gambia on 8th September 1902. All the banking business was exclusively transacted by this
bank for many years as we earlier explained in the introduction to the chapter. The bank
operated well as a pure trading bank, not as a universal, finance or investment bank, though
no other banks existed beside it for many years and the opportunities for other than
distributive trading existed.
This behavior of not venturing beyond trading business was both developed and imported
into the Gambia by the bank. It developed the trading behavior along side the groundnut
marketing, which was the mainstay of the economy. The groundnut marketing
overwhelmed the bank; and with its banking monopoly, it was alone to provide all the
financial services, such as discounting and rediscounting of bills of exchanges, importation
and exportation of currency species, and provision of loans guaranteed by the Government
- 33 -
for the purposes of groundnut marketing. The bank also helped export the profits of the
colonial merchants. Given wide spread network of the bank in other British colonies, the
merchants always found it convenient to deal with the bank for import and export financial
services. The public and the policymakers were all focused on groundnut marketing, and
the distributive trade was profitable; the bank then developed and shaped its behavior
accordingly. Thus, industrial finance, which was a major feature of the banks in the early
stages of industrialization of today’s developed countries never occurred in the Gambia. It
was easy for the bank to concentrate on distributive trade, because that was what it had
mastered in other British colonies before coming to the Gambia. Thus, in part it has
imported the trading behavior into the Gambia.
The Colonial Bank joined the Gambia banking industry in1917; and it closed down after
the Second World War. It was most probably for war financing. The two banks, both of
them were international, experienced the first financial crisis of the Gambia in 1922. The
French silver coin five Franc, then locally known as Dollars (Dallasey) was very popular
with the natives, and most payments were made in it. This forced the Government to make
it a legal tender in 1843. the circulation of the five Franc soared, and in 1880 it formed 85%
of the total money circulation. In 1916, nearly 70% of payments in trade with the natives
were made in French five Franc. The exchange rate was fixed at 3 Gambian Shillings to
one Franc, but for trade purpose it realized 4 Shillings. The two banks in collusion refused
to accept five Franc specie for transfers abroad except at the exchange rate of 4 Shillings;
the currency has become increasingly overvalued. Worse still, in 1917, French Government
prohibited the exportation of the currency from its soil, and in March 1921, the silver coin
was declared non-legal tender. As a result, the Gambia Government banned its importation
into the country. To its embarrassment, the circulation of the currency increased rapidly
forcing the Government to hold over ₤70000 in French five Franc earning no interest and of
no use for transfer. In 1922, the colonial agents declared the five Franc non-legal tender,
and demonetized it with a cost of ₤200000 with 4% annual interest rate. This was the first
time the Gambia Government went into a debt. The banks also shouldered some cost in the
nature of Franc currency holdings. In fact, one could assume that the closure of the
Colonial Bank was partly contributed by this currency crisis. It was established in the same
year the French Government banned the exportation of its silver coin; thus, the bank could
- 34 -
have held huge Franc specie as reserves, which then suddenly proved of no use. The other
theory for the collapse of the Colonial Bank could be that knowing the French five Franc
was overvalued, and knowing that its exportation from France would be banned, the
colonial agents established the bank to start collecting the overvalued currency from the
circulation. Thus, they were embarrassed when the circulation in fact increased, forcing
them to just declare it non-legal tender and asked the public to turn in the currency; and
after the process was complete the bank became of no use.
After the collapse of the Colonial Bank, the Bank of British West Africa continued to
operate lonely in the Gambia. However, its trading behavior gave no help to the
Government’s ambitious development works in agriculture and infrastructure. Financing of
agricultural and infrastructural works was always a trouble. Hoping to give some financial
impetus to the agricultural sector, the groundnut marketing societies called “ co-operatives”
established a Central Co-operative banking and Marketing Union that handled the banking
and marketing business of the societies. However, it depended heavily on loans and
government-guaranteed loans from the Bank of British West Africa. It remained for many
years as one block of loan pyramid with the Bank of British West Africa on the top of the
pyramid. The bank was then phased out, and in early 1980’s it was enjoying only
conditional license. The development financing did not then take off. Wishing it to take
off, the Government and its extensions in the entities of the Gambia Co-operative Union
and the Gambia Produce Marketing Board came together and established in 1972 the
Gambia Commercial and Development Bank. It existed approximately for twenty years.
This period was also one of the booming groundnut seasons, the production averaged
99530 tons from 1972 to 1987, the average highest of all the preceding periods. In this
period, the Gambia Produce Marketing Board, one of the founders of the bank, was making
the highest exploitative profits of all the periods. Its selling price ( export price) was two to
four times the producer price the farmers were receiving. This gave the bank an opportunity
to amass huge deposits for the groundnut co-operative societies, the Gambia Produce
Marketing Board and the Government. However, it built its assets around the marketing of
the groundnut and ignored the production, which was the origin of more than 70% of the
money circulation. Thus, when the groundnut production collapsed, the bank had difficulty
recovering its assets from the groundnut marketing agents. It was the largest bank with 44%
- 35 -
of total bank liabilities before going finally out of business in 1992, its assets were hardly
found in development works but marketing businesses. The Meridian International Bank
took it over, but itself could not survive long. Then came the International Bank for West
Africa. It opened in 1983 in the Gambia, and exited the market before 1992.
It is difficult for the banks to venture outside the traditions of distributive trading to other
areas such as industrial and infrastructural developments. It seems that banks in the Gambia
are followers to the economic activities; they hardly seek out entrepreneurs much more
develop them. Entrepreneurs cannot expect the banks to lead; in fact, they have to wait long
before they can see any bank following; that is, banks are lazy lagers. There could be
reasons for this; one is the low rate of savings in the Gambia and the other is the
inefficiency of the banking industry.
The Gambia has constantly experienced adverse current account balances, from 1935 until
today, the current account has in most cases been deficit. Though the deficit was in some
cases sustained by the re-export trade, its implication for the savings activity in the country
is not encouraging. It always puts pressure on the economy to draw on its wealth and
accumulated savings to sustain the deficit. Because the capital market is not developed,
wealth and accumulated savings are often drawn to finance excess of imported goods and
services. Thus banks are constantly under-funded. The figure 2.6 illustrates the behavior of
the banks’ sources of funds over time.
- 36 -
Figure 2.6
Sour
ces
of F
unds
010203040506070
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
% of total funds
Dem
and
depo
sit
time
and
savi
ngs
depo
sits
Fore
ign
liabi
litie
s
Gov
ernm
ent d
epos
itsC
redi
t fro
m m
onet
ary
auth
oriti
esC
apita
l
- 37 -
Figure 2.6 (continued)
Sour
ces
of F
unds
010203040506070
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
% of total funds
Dem
and
depo
sit
time
and
savi
ngs
depo
sits
Fore
ign
liabi
litie
sG
over
nmen
t dep
osits
- 38 -
Figure 2.6 (continued)
Sour
ces
of F
unds
010203040506070
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
% of total funds
Cre
dit f
rom
mon
etar
y au
thor
ities
Cap
ital
- 39 -
As I said earlier, the local people know little about the banks until very recently. In the
sixties, the banks depended on foreign deposits for funds. From 1964 to 1969, the banks
had zero capital, Government deposits, credits from monetary authorities and foreign
liabilities were the fund sources of the banks. In 1965, over 60 per cent of the bank
liabilities came from foreigners. It fell steadily thereafter, it became 20 per cent by the close
of 1970 and zero by 1972. This decline of foreign liabilities was compensated by capital
injection and borrowing from the monetary authorities. From 1969, banks started injecting
and developing capital, the capital remained below 10 per cent of the total liabilities until
1977, the time it hit 14 per cent. It fell to 5 per cent by 1985, and then it was raised to
slightly above 10 per cent to fulfill the Basel Capital Adequacy requirements. Credit from
monetary authorities remained a strong source of funds for the banks from 1969 to 1986. It
constituted 62 per cent of the funds in 1973 and 47 per cent in 1984. The periods up to
1986, banks obtained their funds largely from foreigners, Government and monetary
authorities but with a decreasing rate of change. Meanwhile, little funds were raised from
the private sector but with an increasing rate of change. The nature of liabilities banks hold
influences the nature of assets they hold. The low level of private funds might not only be
due to the lack of sensitization of the private sector, it could also be due to fact that before
1986 the Gambia financial market was distorted. The exchange rates were pegged to the
UK pound sterling, and over time the Dalasi was overvalued; and the interest rates were not
determined by market forces; the banking operations were also suppressed. Because credits
from monetary authorities dominated the funds of the banks, the Government and its
monetary advisers in the central bank influenced the allocations of the bank resources. The
overvalued currency has also discouraged the private sector and the foreigners from holding
deposits in the Dalasi; as a result, banks were in part forced to borrow from the authorities.
After 1986, the Government lifted sanctions on many market forces, the currency was
floated and interest rates were free. This correction resulted in more than 52 per cent
depreciation of the Dalasi against the US Dollars, and the deposit rate jumped from 9.75
per cent to 16.13 per cent. The funds from the private sector in the form of time and savings
and demand deposits increased, while Government deposits and credit from the monetary
authorities became zero. Time and savings deposits increased from 28 per cent in 1986 to
48 per cent in 1989, an increase of more than 200 per cent in six years. It became the
- 40 -
dominant fund source thereafter accounting for 64 per cent of the total bank funds in 1994.
The second dominant source was demand deposits, the two sources together constituted 96
per cent of the funds in 1994. This trend was tampered with afterwards, savings and
demand deposits declined to 40 and 22 per cents respectively in 1999. The demand deposits
edged up a bit, while savings kept falling. The banks responded by accumulating foreign
assets and injecting capital to boost their fund base. The decline of private funds was a
rational reaction to what was happening to the economy. Deposits with nearly constant
deposit interest rates can be maintained only in a stable currency. From 1999 onwards, the
Dalasi has become unstable, and has been depreciating rapidly against major trading
currencies. The real effective exchange rate depreciated more than 31 per cent within three
years from 1999, and the bilateral exchange rate of Dalasis per US Dollar depreciated more
than 66 per cent. The inflation was also on the rise; while deposit rates were crawling.
These factors combined have discouraged the private sector from depositing funds in the
banks. The banking industry has not also exhibited strong performance. Using the “other
items” in the liabilities side of banking balance sheet to proxy the profitability of the banks,
it shows that banks have been struggling hard to posit any good performance in the entire
period of 1964 to 2002. Other items, which can represent foreign exchange activities or
non-banking operation and extraordinary activities have in some cases more than wiped out
the total capital of the banking industry. 11 per cent of the total liabilities in 1977 is the
highest positive value other items have ever accounted for. It always tended to be negative.
In 1985, it posited a negative value of 21.6 per cent of the total liabilities, and remained
negative until 1997 when it stood at only 1.5 per cent; it then edged up to 4 per cent in 1999
before dropping back to negative. From 1992 to 2002, it averaged at a negative value of
3.12 per cent of the total liabilities. This does not tell well about the performance of the
banks in the country. Figure 2.7 below illustrates the behavior of the account of “other
items” on the liabilities’ side of the commercial banks’ balance sheet:
- 41 -
Figure 2.7
Loss
or G
ain
-25
-20
-15
-10-5051015
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
percentage of total liabilities
Loss
or G
ain
- 42 -
Another factor is that banks have not changed from their old behavior of maintaining high
spreads between the lending rates and the deposit rates, the spread averaged 52 of the
lending rate; in some instances, the lending interest rate is twice the deposit interest rate, as
we can see in figure 2.8 below. This has been a tradition of the banks, which forced the
economy to price everything twice its cost. This is rampant in the Gambian economy; it has
actually been developed by the banks, the monetary authorities and the Government
finance.
- 43 -
Figure 2.8
The other factor is the public lack of knowledge about the banking system. Beyond the 40
kilometers around the capital Banjul, people know little or none about the banks.
Countrywide, we do not have too many banks; the bank density was 0.18 bank per ten
thousand people in 1985, it dropped to 0.15 in 2002. It is only Greater Banjul area that
inte
rest
rate
s
05101520253035
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
- 44 -
enjoyed moderate density of around 0.98. Thus, the number of bank branches can still be
increased to increase competition and force the banks to go up country. We do not need
many banks, but few banks that are well branched and well diversified. This will boost the
degree of bank stability, minimize bank failures and prevent banking panics. Unless these
factors are addressed, the domestic savings and deposits in the banks will continue to
constitute low source of the bank funds, the banks will be under-funded, the money will be
expensive and development projects will hardly take off.
The other reason why development projects do not qualify for the bank funds is due to the
inefficiency of the banks. The commercial banks have been reducing their claims on the
private sector and replacing it with foreign assets and claims on Government and official
entities. From 1964 to 1973, commercial banks’ claims on the private sector accounted for
more than 60 per cent of the bank total assets. It kept fluctuating between 60 and 40 per
cent from 1973 to 1996; then it remained below 40 per cent . This is because the
Government and its extensions compete with private sector for the under-funded bank
funds. In 1975, the banks withheld funds from the private sector and gave it to the official
entities. In 1978, the private sector was given and the official entities were denied, and the
opposite occurred in 1983. That is, the fund is not sufficient for the two sectors, if one gets,
the other will be denied. Figure 2.9 illustrates the distribution of the commercial banks’
assets.
- 45 -
Figure 2.9
Dis
trib
utio
n of
Ban
ks' A
sset
s
020406080100
120 19
64
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
% of total assets
Res
erve
Fore
ign
asse
tscl
aim
on
Gov
tC
alim
s on
Offi
cial
ent
Cla
im o
n pr
ivat
e se
ctor
- 46 -
Later, banks became very inefficient and complacent; they found it more convenient to deal
with Government and official entities, where monitoring and researching are less required
than in the private sector where bankers must be well trained and monitoring and
researching are costly. Thus, the bank claims on the official entities rose from 11.12 per
cent of the total bank assets in 1978 to 38 per cent in 1982; the claims on the private sector
dropped from 66 per cent in 1978 to 38 per cent in 1982. From 1992 to 2000, banks wound
up their claims on the official entities and concentrated on the Government and the foreign
assets. By 2001, the banks’ claims on the Government stood at 47 per cent of their total
assets.
In all, the private sector stands to lose. Its savings and deposits have an upward trend, while
its share of the banks’ assets kept declining. It could be that the borrowing behavior of the
Government has crowded out the private sector; or the increasing debt of the Government
has increased the country risk, and as a result the private sector activity declined. The
country risk factor could be deduced from the foreign asset behavior of the banks. Banks
have increased their foreign asset holdings from 0.67 per cent in 2001 to 18 per cent in
2002, the second highest foreign assets holding in the entire period of 1964 to 2002. The
inefficiency to develop entrepreneurs and lead the market has also led to falling claims on
the private sector. The shrinking claims indicate that the economy is nurturing inefficiency
in ballooning public debt and increasing foreign assets. This was a pre-independence
phenomenon, when the banks exported funds to foreign lands and the Government held its
account in foreign assets, the domestic economy was left to starve. Today, the public debt
has jacked up the country risk, and the transaction costs of the banks have gone up with
lending interest rate almost twice the deposit interest rate; the private sector is left to starve
for funds. With this phenomenon, the financing of development projects especially the
private sector projects may remain a dream.
A micro-view of banks’ claims on the private sector can tell us exactly what happened.
Figure 2.10 below indicates clearly that the major economic sectors have been over time
losing the commercial banks’ support.
- 47 -
Figure 2.10
Ban
ks' C
laim
s on
Maj
or E
cono
mic
Sec
tors
010203040506070
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
% of total bank assets
Agric
ultu
reM
anuf
actu
ring
& Fi
shin
gBu
ildin
g &
Con
stru
ctio
nTr
ansp
ort
Dis
tribu
tive
trade
Tour
ism
- 48 -
The groundnut production which dominates the agricultural sector has been losing
dramatically its share in the total bank assts. Its share has fallen from 66 per cent in 1973 to
4 per cent in 1991. Agriculture is the mainstay of the economy, it employs more than 70 per
cent of the population, but has in recent years almost nothing to do with the banks. Have
the banks found a better economic sector for their money? Looking at figure 2.10, it is
difficult to tell where the banks are placing their money. The other major sectors have
approximately maintained their shares from 1975 to 1995. The distributive trade, which has
the second highest share until 1982, declined only slightly. The individual shares of the
other sectors remained below 10 per cent. That is, banks do not improve any other sector as
they withdraw funds from the agriculture; they are just increasing foreign assets base and
claims on the Government. This investment strategy did not improve the assets of the
banks. The total bank assets as percentage of GDP have been falling from 53 per cent in
1982 to 20 per cent in 1992. This confirms our assertion that banks are inefficient and
incapable of developing entrepreneurs. Thus, we expect to empirically find that economic
activities Granger cause banking activities, contrary to the plausible findings that banking
activity Granger causes the economic activity. The latter occurs in an efficient and
entrepreneurial banking industry, while the former occurs in an inefficient and laging
banking industry.
- 49 -
Figure 2.10 (continued)
bank
s' a
sset
s an
d ba
nks'
cla
ims
on m
ajor
eco
nom
ic o
pera
tions
0102030405060708090100
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
Cla
ims
on m
ajor
priv
ate
sect
ors
as %
of b
ank
asse
tsTo
tal b
ank
aset
s as
% o
f GD
P
- 50 -
2.5 Government Finance and Taxation Underlying all the questions of economic progress and capital development in the Gambia
is the desperate recurrent financial condition of the Government. Up to 1951, the
Government finance was simple but unsatisfactory. The finances were built up, while there
were no corresponding expenditures in the development of the resources of the country.
The staff of the Government were mostly foreigners, who were not trained by the country’s
finances; the communications were rudimentary, the groundnuts, the most important export
produce of the country, could be easily loaded at wharves along the River Gambia onto the
ocean-going vessels; so, there was no acute need to build railways or construct good roads.
Little was also spent on education; since no need arose for educating the masses to cultivate
groundnuts, tap rubber, crack palm kernels, or collect hides and skins, which were the
export goods of the country. When the Government wanted an agricultural expert, it could
cheaply get one seconded from India or China; and when it wanted a vetinary,
administrator, or educator, Nigeria and Sierra Leone were always there to supply. The
whole country could be run by less than one half of the inhabitants of Bathurst (Banjul),
who were in the country the educated West Africans known to Europe.
The Government depended for finances on custom duties, land rents, hut taxes, and
individual and company taxes. Currency Board Profits and interest incomes from Overseas
Government investments also constituted some source of Government revenue. However,
in the face of limited or non corresponding expenditures in the development of the
country’s resources, the revenue sources of Government were bound to dry up. This was
acknowledged by the Governor at the Budget Session in December 1951 as he warned the
Legislative Council saying: “we have now tapped to the full all available sources of
revenue and until we have raised production in this colony we have reached the end of our
taxable capacity”. It was like the story of rubber plants in the Gambia. Rubber was obtained
from wild landolphia vines, and was not cultivated. The vines were tapped all the year
around, a most uneconomic method and one which has helped to kill them off. By 1920, the
rubber disappeared from the export list of the Gambia. Similarly, the Government
continued to tap the country’s resources without cultivating them, which has no doubt
helped to drain up the taxable capacity of the country.
- 51 -
By 1950, groundnut remained the only significant export of the country, and over ninety
per cent of the population depended on it for cash to finance their expenditures. Raising
export duties would eat away the producer prices and render the ninety percentage of the
population poorer than before, and raising import duties would raise the costs of living and
the quality of living standards would be compromised. In fact, in 1958, the Government
had to reduce the export duty on the groundnuts by ₤1.5 per ton and pay a subsidy of ₤2 per
ton to the Gambia Oilseeds marketing Board to enable it to pay a fixed producer price of
₤27 per ton. The government revenue from groundnut export duty later dwindled to nothing
as the World prices of groundnuts and groundnut oil kept falling, and the Government was
called upon every succeeding year to come and save the Gambia Oilseeds Marketing
Board, the sole purchaser of the groundnuts and to subsidize the producer price. Groundnut
cultivation was the only industry of the economy, and now no revenue could be tapped
from it without holding the farmers at subsistence level or letting them go below it. There
were many factors responsible for the collapse of this industry; these factors will be
discussed under the section Trade and Industry.
- 52 -
Figure 2.11
cust
oms
cont
ribut
ion
to G
over
nmen
t Fin
ance
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9 1903
19051907190919111913191519171919192119231925192719291931193319351937194019421944194619481950195219541956195819601962
percentage
- 53 -
The government then turned to other taxes such as individual and company taxes for
finances. Individual income tax was introduced in 1940, but did not fetch much revenue;
little human capital was developed or generating income. Meanwhile, the companies were
enfant and required some tax haven and exemption from some set up costs as licenses and
fees; levying increasing taxation or license costs and fees would cause them to go out of
business or leave the domestic market. Thus, these sources of Government revenues were
not sustainable. The other unstable revenue source was import duties. Import duties depend
on the volume of imports, which in turn depends on the domestic income. With the collapse
of groundnut market, which was the source of income for the ninety per cent of the
population, the import dues were certain to fall. For some time, this import market
flourished due to demands in the neighbouring countries. The Gambia with no plans of
developing domestic industries, its import tariffs were relatively cheaper than its
neighbours, who were trying to do something for local industries. The neighbours also had
one import source, France; while the Gambia was supplied by more than six western and
Asian countries. The Gambian import was swollen by the demands in the neighbouring
territories, as traders responded by re-exporting or smuggling the excess imports to the
neighbours. The Government finance was precarious and at the mercy of the neighbours;
the ordinary revenue hardly met the recurrent expenditures, and that any new capital work,
amenities and services required outside help. By 1958, the Government has already given
up and separated the ordinary budget from development budget, and the later was almost
entirely funded by grants and loans.
- 54 -
Figure 2.12
Bud
get d
efic
it/su
rplu
s
-800
000
-600
000
-400
000
-200
0000
2000
00
4000
00 19001902190419061908191019121914191619181920192219241926192819301932193419361938194119431945194719491951195319551957195919611963
Pounds
- 55 -
2.6 Trade and Industry
The economic activities in the Gambia up to very recently were centred on the production
and marketing of groundnuts. No minerals of commercial value existed; no other crops
could get the attention of the people because it paid them higher to cultivate groundnuts
than other crops. Thus, groundnuts remained the sole important crop and the only industry
accounting for over 80% of the national export for over 80 years. Groundnut trade was
introduced into the Gambia in 1835 in consequence of demand for nuts in North America
and England. Since then until 1950’s, the trade prospered progressively; the producer price
kept rising along the increased production. Farmers were receiving ₤4 to ₤5 per acre in
1908’s. This gain progressed well until it hit the peak at ₤40 per ton in 1951. Thereafter,
“trade” as locally referred to broke no good news to the farmers. In addition to external
factors, population and finance played important roles in the rise and fall of the groundnut
industry; and the fall of the groundnut industry has contributed significantly to the
stagnancy and retrogress of the Gambian economy.
- 56 -
Figure 2.13
Gro
undn
ut s
hare
in th
e ex
port
020406080100
120 1843
18631883190319051907190919111913191519171919192219241926192819301932193419361938194719491951195319551957195919611963
Gro
undn
ut s
hare
in th
e ex
port
- 57 -
Groundnut was demanded internationally for its oil; and until 1960, the Gambia exported
raw groundnuts to be processed in the importers’ lands. After 1960, the Gambia started
processing groundnuts and exporting its oils. This project coincided with the advent of new
sources of edible oils, whose inputs did not originate from the Gambia. As varieties of
edible oils overflow the world market, the local groundnut oil refinery faced a difficulty in
giving good prices to the farmers for their groundnuts and the same time sell its oils at
profitable prices in the world market. It is a small industry and its output is too small to
influence the world prices of the edible oils. Thus, the local refinery loses out any time a
better and cheaper edible oil appears in the market, and the only way for it to make profit is
to reduce the purchase prices, which means impoverishing the farmers every succeeding
trade season. From 1957, the local refinery could purchase groundnuts only at a subsidized
price or at a reduced export tax. This was the most important external factor responsible for
the fall of the groundnut industry. There are two local factors, population and finance.
Population
There are many other crops such as millet, African Koos, and maize, but they are not cash
crops, they are for local consumption. But they compete with the cash-crop groundnut for
the land and labour. The high market prices of groundnuts induced the farmers to cultivate
more groundnuts, because groundnuts receipts could get them local foodstuffs or imported
ones. The farmers concentrated on groundnut cultivation and less attention was given to
food crops. In 1904, when the governor asked the people to cultivate foodstuffs and realize
self-sufficiency in food, the chief of Tambasansang, Farli Cora responded: “I am ready and
willing to take up these things you tell me about, but which of them pays best? As my
people are so few and they could not deal with them all”. Well, it was only the groundnut
that could pay farmers the colonial monies. The colonial monies, which were British silver
coins, French five francs, and notes of Bank of England, were imported into the country
either for paying the colonial government officers or for paying for groundnut produce.
Only in the latter, did the ninety per cent of population, who were the farmers, could see the
colonial monies in their hands. Thus, the colonial monies flowed high during the trade
season and their circulation fell very low afterwards forcing many farmers to resort to
barter trade, or invoke traditional pre-colonial monies. To get more colonial monies was to
- 58 -
produce more groundnuts. With no capital and more lands left uncleared, to increase the
production the farmers had to increase the labour. Any additional labour to the land would
increase the output. The more numerous the number of the male members of the compound
the more groundnuts would be harvested by that compound. As a result, farmers planned
for large compounds; and seasonal farmers from the neighbouring territories were welcome
to the Gambia to cultivate and sell groundnuts in the Gambia. Over time, small families
grew into large compounds and more seasonal farmers preferred to settle and grow
groundnuts in the Gambia than going back and coming every groundnut planting season.
Thus high amount of labour was added to the land and the production swollen. Then, the
diminishing returns quickly set in due to the fact that land is fixed and increasing farming
population would definitely after a point cause the output to fall. Furthermore, for high
yields per acre the farmer had to alternate the cultivation of millet and groundnut on that
acre for five years; after that the acre had to be uncultivated for 15 years for it to revert to
bush and for the topsoil to have a complete regeneration. This makes it 20 years of
regeneration cycle. With small farming population, this regeneration could complete its
cycle; meanwhile, increasing farming population reduces the regeneration cycle. This in
fact caused the diminishing returns to quickly set in in the groundnut industry. Also, the
large compound initially boosted the groundnut production because of its large members
and the harmony of its structure. All the members of the compound worked under one
member, who was the head and financier of the compound. The members together could
clear more lands and produce more output than doing that individually. From 1960’s, the
large compound structure started giving way to small family units. This caused changes in
the farming methods; instead of all members, even subfamily units, working under one
head, now each member or subfamily unit worked alone, and the farming land of the then
large compound got divided into small plots. The farming plots became smaller and smaller
as the compound structures broke up or sub compounds disintegrated. This disintegration of
compound structures contributed to the fall of groundnut output.
- 59 -
Finance
Financial intermediaries, whether via government finance or traditional moneylenders or
the banks, all profiteered from the groundnut farmers. Government reaped a lot of export
duties on groundnuts; in 1951, the only way for the Government to balance its budget was
to increase export tax on groundnuts. But little did the Government do for the farmers
financially until the industry slipped into a crisis of no-way out. The groundnut season runs
from June to March of the following year: the seed is planted in June to July, the crop is
harvested in October to November, and the nuts are threshed and marketed from December
to March. Depending only on the earnings of the groundnuts, which are cash payments in
the months of December and February, the farmers run all the cash transactions (no credit
for farmers) throughout the off-season trade. Also the same cash receipts of the groundnuts
are to be used to finance the working capital of the groundnut production. That is, farmers
are producers and financiers. No banks around to keep the money; thus, the farmers keep
the cash payments under their pillows and praying mats. In many instances cash balances
from the trade season turn to zero or negative in the very months of threshing and
marketing of groundnuts. The money received during the trade season quickly finds its way
back to Banjul where banks and Government departments are. This occurs through the
merchants, who sell imported products and manufactures to the farmers, or through the
farmers themselves as they travel to Banjul to do their shopping. The Government did not
build money catchments around the farmers, neither the branches of the dead currency
board existed and nor that of the current central bank existed anywhere in the country
except in Banjul. Because modern banks cannot exist without a central bank, it implied that
no branches of commercial banks could also exist anywhere in the country except in
Banjul. This is what exactly happened in the Gambia. Farmers have no choice but to take
the money to Banjul as they come there to shop or buy goods from trade merchants. This
scarcity of money during the off-trade season makes the money very expensive at the time
when the farmers are looking for working capital to start the planting season. Thus, those
farmers who accept loans from moneylenders pay high prices to the extent that some of
them will give all their groundnut earnings as loan repayment to the moneylenders. An
outcome of money scarcity is a barter trade to finance a working capital; for example, a
farmer would accept 50kg of groundnut seeds from a moneylender on an agreement that he
- 60 -
will give the moneylender 100kg groundnut at the trade season as repayments for the
principal and the interest. Also, the farmer barters his labour as he works on others’ farms
and accepts payments in goods usually foodstuffs. What have the Government done for the
farmers? And what should they have done? We start with the first question.
In 1948, the Government instituted a Co-operative department and passed an ordinance
creating Gambia Oilseeds Marketing Board and gave the Board (GPMB) the exclusive right
to conduct large-scale buying and selling of oilseeds. The department was to help the
farmers run the groundnut production; it supplied small agricultural tools and financing
facilities to the farmers. In the first experiment, in 1956, the Government provided ₤1200 at
5 per cent for subsistence loans to farmers, it was fully repaid. In the following year ₤4200
was advanced to the farmers. By 1964, Co-operative groundnut societies covered the whole
nation. A lot of funds were mobilized from these societies; by 1960, farmers’ fund stood at
₤905000 and kept rising; it was not invested in farmers but abroad in the UK assets. Later,
when the groundnut prices started falling, farmers’ funds were withdrawn to subsidize.
Little was actually invested in the farmers. For many years the governments imposed taxes
on the produce and fixed the buying price, and then utilized the surplus retained by GPMB.
This surplus was known as the farmers’ fund “, it was used not to stabilize prices as
originally intended, but instead to set up the Fish Marketing Corporation and the Livestock
Marketing Board, both of which were put into liquidation only two years after their
establishment; in effect wasting the groundnut producers’ forced savings”7. Funds were
originally intended for setting up and financing rural projects that would have direct impact
on the farming community; but few of these projects such as mangrove clearance, pest
destruction, improving village water supplies and ferries, saw light. The Co-operative
Central banking Union afterwards could only purchase groundnuts by borrowing. In a
consequence, a pyramid of debt structure was created, where the currency board, later the
central bank, or the Bank of British West Africa would lend to the Government, and the
Government would re-lend the funds to the Co-operative, which in turn would re-lend to
the farmers or use the funds to purchase the produce. The illiquid plight of the farmers got
aggravated as the debt pyramid increased the transaction costs of the funds. As the situation
7 The marketing of Foodstuffs in the Gambia 1400 – 1980, Hazel R. Barrett, Gower Publishing Company Ltd, 1988, p.44
- 61 -
worsened every succeeding year, good farmers dropped out of the scheme and an adverse
selection of farmers swollen the scheme leading to the eventual failure of the co-operative
department.
What should the Government have done? First, there should have been branches of the
central bank in every provincial capital to facilitate the upkeep of money around the
farmers and prevent it from draining down to Banjul. This would encourage the expansion
of commercial banks and other financial institutions in the up-country. Second, the co-
operative board of the Government should have been in a position to buy and sell the
farmers’ produce both before and after the harvest. Farmers should have been able and
encouraged by the board to enter into futures contracts and sell their produce to the board to
ease the illiquidity problem they faced in the planting and harvesting periods. Finally, not
all farmers would prefer to run their own farms, they would rather prefer to sell their labour
during the farming season. These people should have been able to do so either by being
paid to work on other private farms or on state run farms. These three mechanisms could
have resolved the financing problems faced by the farmers.
Balance of Trade
One will not be wrong if one says groundnut is the only trade of the Gambia; and when it
started collapsing the whole economy started collapsing. High market prices for groundnuts
year after year, limited the growing of other crops and gave high returns to farmers and
traders that they found hard to take up other industries seriously. Furthermore, until 1951,
staple foods, building materials and small items were exempt from import duties. Thus,
farmers could afford the imported items with their groundnut money. But farmers could not
spend more than the export values of groundnuts; so, what has accounted for trade deficit
since 1941. There are five factors that account for sustainable trade deficit in the Gambia;
and credits of these factors do not represent significant capital ploughing into the Gambia,
and worse still most of them go unrecorded.
The first of these factors is drawing on the wealth and accumulated savings by the farmers.
During the hay days of the groundnut industry, some farmers built savings and wealth in
businesses and cattle, and at the same time they planned for large expenditures in the form
of large families. When the industry started collapsing and current earnings proved
- 62 -
insufficient to finance the recurrent expenditures, farmers drew on their wealth and
accumulated savings to balance the budgets. Over time, the wealth and the savings dried
up, and the declining groundnut earnings could not be saved or reinvested, the farming
community broke up and the able members started travelling abroad. Thus, the second
factor for sustained trade deficit is remittances by Gambian nationals. From 1950’s, the
time groundnut industry began collapsing, until today, large remittances have come to the
Gambia unrecorded; and these remittances constitute the most significant income for most
of the former farming families. As a result, the families back home can still afford the
imported goods including the foodstuffs. At present, any capital expenditure by former
farming community is mostly financed by remittances. The third factor is the expenditures
by the Government departments and organizations. Pen and paper, which are essential
operating expenses of the Government, are not even assembled in the Gambia much more
manufactured. You can say similar things for many other operating expenses that constitute
the bulk of ordinary operating expenses of the Government. Expenditures by non-
governmental organizations (NGO’s) also constitute a factor for the sustained trade deficit.
Most of the schools in the Gambia are run by the NGO’s. Spending by these NGO’s is
financed in part by overseas donors and sponsors, and the funds normally come unrecorded
and then get spent on the imported goods contributing to the adverse trade balance. Finally,
demands in Senegal and other neighbouring territories for goods imported into the Gambia
plays an important part in increasing trade deficit. Re-exports to Senegal are not recorded,
only small re-exports to other neighbours are captured in the trade account. The current
account balance as percentage of the GDP has been fluctuating wildly over the 1964 -2002
period. The surplus has in no time reached 20 per cent of the GDP, and it was around 10
per cent of the GDP only for the period 1915 to 1933. In many years deficit has been the
mark of the Gambia current account. The deficit has stood around 10 per cent of the GDP
for the most part of the period 1964 – 2002, and exceeded 20 per cent of the GDP in six
instances. It stood above 30 per cent in 1979 and above 40 per cent in 1982. Thus, the
economy has in general bad international trade activities.
- 63 -
Figure 2.14
Cur
rent
Acc
ount
Act
ivity
-50,
00
-40,
00
-30,
00
-20,
00
-10,
00
0,00
10,0
0
20,0
0
30,0
0 1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1946
1949
1952
1955
1958
1961
1964
1967
1970
1973
1976
1979
1982
1985
1988
1991
1994
1997
2000
Cur
rent
acc
ount
as
% o
f GD
P
- 64 -
The gross domestic product gives us a summary indicator of the economic health of the
Gambia. No compiled GDP data do exist for the period 1900 – 1963, I instead estimate it
using the available data on the government expenditures, exports, imports and the
groundnut output. I proxy the private investments by the groundnut export value, because
groundnut was the most active private industry, around 90 per cent of the population
engages in its production. They produce it for the export. I proxy the private consumption
by the groundnut output. The percentage they locally consume out of the raw groundnut
was very small; they had to sell almost all the groundnut output to buy the consumption
including the processed groundnut. But since the accurate data on the groundnut output was
available only for the volume brought for trade, which was equal to the groundnut export
value, I use the same groundnut export value for the private consumption. This gives us the
GDP for the period 1900 -1963 as follows:
GDP = Government Expenditure + 2 Groundnut export Value + Net Export
For the period 1964 – 2002 , I use the available GDP data in the secondary sources. I then
take the log of this GDP and that gives us figure 2.15 on the next page. The GDP found an
increasing path only from 1948. It was growing at around 1.30 per cent annually around
1948, then at around 1.5 per cent in 1953. It took it 20 years from 1953 to 1972 to reach an
annual growth rate of 2.0 per cent , and it was able to double its 1953 growth rate only after
33 years. It went to an annual growth rate of 3.5 per cent in 1993, and it was growing at
around 4 per cent in 2002. We measure the performance of the Gambia economy by
calculating the number of years the economy takes to achieve an additional 0.5 per cent in
growth, we find that the best time was from 1972 to 1977, it took the economy 6 years to
achieve an additional 0.5 per cent in growth; while the worst period was 1917 – 1947,
during this period the economy stagnated or declined, and the annual growth rate stood
around or below 1 per cent. The second worst period was 1953 – 1972, it took the economy
20 years to achieve 0.5 per cent in growth. It took also 15 years from 1903 to 1917 to
achieve an increment of 0.5 per cent. The recent years have not either been promising, we
are now 10 years from 1993 to 2002, and the economy has not been able to achieve an
increment of 0.5 per cent. This is the Gambian economy, it is stagnant or very slow moving
because of desperate sole agricultural industry and precarious government expenditure.
- 65 -
Figure 2.15
However, I can give some major reasons for the slow performance of the economy over
the periods, 1900 – 1917, 1917 – 1947, 1947 – 1953, 1953 – 1972, 1972 – 1977, 1977 –
1986, 1986 – 1993, and 1993 – 2002. Each period, except 1917 – 1947 and 1993 – 2002,
Gro
ss D
omes
tic P
rodu
ct
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
4,00
4,50 1900
1903190619091912191519181921192419271930193319361946194919521955195819611964196719701973197619791982198519881991199419972000
log
GDP
- 66 -
measures the number of years the economy took to achieve an increment of 0.5 per cent in
growth; and I call this number of years the speed of growth. The first period, 1900 – 1917,
was the monetization period of the Gambian economy and its ascendancy to the capitalist
world; and the economy was experimenting different crop productions for its colonialists
until it settled on the groundnut, which was exported to factories in North America and
England. The industry then faced logistic problems, the population was under 200,000 and
was not willing to learn new methods of production; the total land of the country is about
11,000 squared kilometres and less than 20 % of it could sustain the groundnut produce.
The storage and the transport of the crop within the country were also inadequate and
rudimentary resulting sometimes in infected nuts. Thus, the supply of the nuts was
logistically constrained, and this forced its demanders in America to switch to other
alternative sources in their local economies. Then Britain and France became the sole
demanders for the Gambia groundnut; French Franc and British pounds started flowing into
the Gambia economy, the trade seasons flourished and the farmers had surplus funds.
Production of other crops was almost abandoned, and farmers concentrated on producing
cash-crop groundnut, the neighbours were attracted into the country and the farming
population quickly swollen, and all the untapped arable lands were cleared. This boom of
the first period was actually going to burst due to its constrained logistics of swollen
population, repeated use of unimproved lands and failure of the farmers to adopt new
methods of production. Thus, the second period 1917 – 1947 was a natural burst of the
boom of the first period, but the burst or the recession was prolonged due to the additional
man-made events. This period witnessed the first bankruptcy of the Gambia government
when the French government stopped the importation of the French franc by 1917 and
declared it non-legal tender by 1922, the French franc was then constituting around 70 % of
the money circulation in the Gambia. In other words, 70% of the Gambia means of
payment became null and void; the farmers, the business communities and the government
were all 70% bankrupt, the economy was later assisted by a loan from the West African
Currency Board. The other man-made event was the world wars that almost zeroed the
groundnut exports to France and Britain; and worse still the British established trade boards
in its colonies including the Gambia to stop the export supplies from reaching its enemies
who were willing to pay more than the British. This period lasted for 31 years, and the
- 67 -
economy could hardly grow above the rate of the first period. It would have been shorter
had it not been the man-made events. The economy entered into the third period from 1947
to 1953 with a good start, the world wars ended, the American Reconstruction Aid to
France, Britain and Europe in general was causing positive ripples in the Gambia economy,
the groundnut export lines were fully restored back, and the British established a colonial
development corporation to help rebuild the Gambia economy; in 7 years the economy
achieved an increment of 0.5 per cent in growth. But this speed could not be increased or
even maintained as the economy entered into the fourth period, 1953 – 1972. This period
witnessed independence struggles across Africa, and France and Britain started rethinking
their cooperation with or dependency on the colonies in Africa, the rethinking deepened
particularly when the first independences were sort of antagonistic and unfriendly divorce.
The projects of the colonial development corporation in the Gambia, such as rice farms and
poultry farm were left to perish. The French and British factories reduced their dependency
on the Gambia groundnut; and worse still the trade boards established during the wars to
stop supplies from reaching the British enemies were then used as an attempt to tighten the
colonial grip on the colonies as the independence struggles heightened. This cost the
Gambia economy 20 years to achieve an increment of 0.5 per cent in growth. The following
period, 1972 – 1977, was the best period for the Gambia economy. In this period,
confidence was restored into the former colony, the economy enjoyed the highest
groundnut production and export value of all the past periods. It is in this period that the
Central Bank of the Gambia had to revalue the Gambia currency peg to the British pound
because the latter had depreciated significantly against the Dalasi. Also, there was thy-
neighbour effect, as the biggest economy in the West Africa, Nigeria, was reaping high oil
revenues due to its non-participation in the OPEC oil embargo in 1970’s. This effect has
spilled over to the Gambia, and trade flourished between the two countries and many
Gambians sought works in Nigeria. Thus, the period had the highest speed, in 6 years the
economy achieved an increment of 0.5 per cent in growth. This speed then reduced in the
sixth period probably due to three events, the shocks of the IMF-structural adjustments, the
shocks of the financial liberalization in the Gambia and the shocks of the attempted
overthrow of the Government. The economy slowed from six years in the last period to 10
years in this period to achieve an increment of 0.5 per cent. It increased to eight years in the
- 68 -
seventh period, 1986 – 1993, the fruits of the financial liberalization were reaped in this
period. The attempted overthrow of the government strengthened the bilateral relationship
between the Gambia and Senegal because the latter restored back the Gambia government,
and confederation was then signed between the two countries. This helped the Gambia
expand its re-export routes across the borders with Senegal. The Gambia sea port became
one of the busiest ports in the region as countries neighbouring Senegal and Senegal itself
increased their demand for the goods transited via the Gambia sea port. But recently, 1993
– 2002, the growth has slowed, the confederation was abandoned, the Senegalese are no
longer overlooking the border crossing of the goods from the Gambia. The groundnut
industry could be declared dead, and the labour from the agriculture streamed into the
urban places increasing the urban unemployment and pressure on the facilities. The
overthrow of the government and the consequent halt of tourism further depressed the
economy in this period. The financial mismanagement at the central bank and the desperate
government finance are also to blame for the low economic performance in the recent
period.
2.7 Conclusions
The Currency Board System succeeded in steering a better economy for the Gambia than
the Central Banking System. It enjoyed low inflation, price and exchange rate stability and
steady economic performance. It did so until the Government amended the Currency Act;
an amendment that effectively disabled the Currency Board and transformed it to more like
a Government commercial and development bank than even a central bank. The Central
bank came at the time when the sight liabilities of the Currency Board far exceeded its
foreign assets; the foreign asset holdings dwindled to an unmanageable level; the Board had
to be effectively declared bankrupt and taken over by a central bank. The Central Bank
continued the benevolent services effected by the Currency Act Amendment, and it became
blind to the market information as prices were controlled, interest rates fixed and currency
pegged. The Government found a good source of financing its deficit not only through
borrowing from the Central Bank, but also through borrowing from the captive commercial
banks, where it could borrow at below market interest rates. Over time, the economy
became saddled with inefficiencies due to suppressed market information, Government
- 69 -
deficit financing option and neglected agricultural industry. The Central Bank also lacked
innovations in the payment system so much so that a teacher in a far end of the country had
to travel all along to Banjul to receive his salary, and a paymaster had to load motor cars
with Dalasis to go around the country paying his subjects. The Central Bank did not branch
out, and nor did the commercial banks; thus the capitalist property got concentrated in
Banjul and its surrounds triggering the urbanization problems. This thing became worse as
the neglected agricultural industry crumbled. The produce of the agricultural industry was
exported raw from the farm fields to the overseas. No processing industry was installed
early to harness the surplus labour that would be produced by the agricultural industry. As
we said earlier in section 2.5, in the beginning there was a labour shortage in the
agricultural industry, but the bumper harvests, foreign exchange earnings of the groundnut
export and wide available untapped land acreages sustained increasing farming population.
This process later produced surplus labour; with no other industries to employ them, the
surplus had to be redeployed back to the farm resulting in low farm productivity, low per
capita farm output, low per capita food consumption and stretched or unbalanced family
budgets. The groundnut industry did produce enough savings and foreign exchanges for the
economy to kick start the manufacturing stage. But unfortunately the savings forced out of
the farmers’ purchasing power were wasted on plants set up but only to be liquidated after
two years, or wasted on subsidizing unsustainable rice projects. The savings and foreign
exchanges of the groundnut produce went almost into all sectors of the economy; early
commercial banks were purposely set up to finance the merchant business, which was
dominated by groundnut trade; the Government deficit and the Central Bank currency peg
were essentially financed by the foreign exchange earnings and savings of the groundnut
industry. The funds were not used to cultivate or invest in the farmers, and thus the industry
was doomed to collapse. Lack of other industries that should have effectively used the
groundnut surplus labour and foreign exchanges to produce the basic consumptions and
capital goods for the farmers had caused the economy to be entirely dependent on the
imports. Under these situations, suppressed market information and adverse consequences
of the neglected groundnut industry, the Government and the Central Bank stroke a reform
in 1985/1986. The price controls were lifted, the Government attempted a minimal
government by setting up some parastatals; the interest rates and exchange rate were let
- 70 -
afloat. Banks were then able to attract more savings and foreign liabilities than before. This
increased bank liability did not sufficiently translate to increased lending to the private
sectors or long term commitments. In fact, the banks’ commitments in the major economic
sectors have been falling. Increased interest rates and unstable prices and exchange rates
have increased the transaction costs for both domestic borrowers and lenders. Thus, the
banks had to transfer the increased liabilities into Government and Central Bank securities.
The Central Bank also in an attempt to curb the currency and price falls has been amassing
foreign assets and at the same time sterilizing the national currency through increased
Central Bank bills, which are often borrowed at zero costs to finance the government
deficits.
This is a summary of the Gambia economy. The solution is economic governance and fiscal
discipline are required on the part of the Government. The Central Bank must be allowed to
pursue only one single objective, which must be price stability. It should be responsible to
the Parliament and periodically report its performance and actions taken to the Parliament.
The Central Bank should not act as a lender of last resort, nor a supervisory entity for the
commercial banks to avoid conflict of interest. The commercial banks must forge among
themselves to set up a lender of last resort and set out the requirements and functions of that
body. The Government should borrow from the voluntary lenders in domestic currency
market. The Central Bank should periodically report to the Parliament and public the
structures of the commercial bank liabilities and assets and highlight the commercial bank
commitments in the major economic sectors. The research department of the Central Bank
should have, among others, a specialised unit for each and every major economic sector to
continuously bring to the light the strengths, opportunities and weaknesses of every sector.
Finally, the Central Bank should branch out to increase commercial bank branching and the
safety and efficiency of the payment system and nationwide coverage of the financial
services.
- 71 -
3 Survey of Theory and Evidence
3.1 Introduction
This section discusses the theories for the emergence of banks and the empirical evidence
on the bank-growth or finance-growth relation. Information economics holds that banks
emerge as a result of information asymmetry between the deficit agents and the surplus
agents. The banks efficiently collect funds from multiple surplus agents to finance one or
more deficit agents, who on their own will find it impossible to do so. In this way the
transactions costs for both surplus agents and the deficit agents are minimized. The
implication of this theory is that the banks enable or facilitate the economic transactions;
and if a more efficient intermediation can be found, the role of banks will be eroded if not
totally eliminated. The other theory, which we call here evolutionary theory, claims that
bankers were actually big merchants or firms. They evolved from loaning their excess
funds to other merchants and firms, to discounting bills of exchanges and to issuing their
own bills. These merchant bankers relied on their own equity to perform banking
operations. The merchant bankers do not only enable the economic transactions but they
also anticipate them by creating their own bills for the future economic transactions. We
then discuss the four empirical evidence on the finance-growth nexus.
3.2 Existential Theories of Debt- Intermediation
Why does debt – intermediation exist? What factors are responsible for its continuous
existence? We explore in this section two theories that claim to have answers for the above
questions. One is embodied in the information economics; that is informational asymmetry
and transaction costs. This is the most widely held theory; while the other, we term it
evolutionary theory of debt –intermediation. Debt – intermediation institutions, here known
as banks, research potential investments, exert corporate control, manage risk and mobilize
savings. They accept deposits from and lend to large number of agents. They issue
liabilities, such as demand deposits, savings, that are more liquid than their primary asset,
such as government securities and personal loans. They eliminate or reduce the need for
self- financing. Liquidity need may force entrepreneurs to prematurely liquidate their
- 72 -
capital investments; banks intervene to prevent this unnecessary liquidation. In addition to
their main functions of interpersonal and inter-firm resource transfer, they also do inter-
temporal smoothing. “They acquire a buffer of short – term liquid assets when times are
good and run this buffer down when times are hard” ( Allen et al., 1993).
The question to discuss in this section is why the intermediation exists in the first place.
Why the deficit agents do not engage directly with surplus agents and execute debt
contracts and in the process eliminating the role of intermediation. We discuss this under
the following two theories, information economics and evolutionary theory.
3.2.1 Information Economics
Debt contracts are hardly free from problem of informational asymmetry, which are
adverse selection and moral hazards (Bolton et al., 2005). The creditor may know little or
none about the creditworthiness of the borrower or about the safety of the project. This lack
of information on the part of the creditor about the borrower type or about the project type
may cause the creditor to increasingly charge high interest rates for the funds, which
consequently drive away the safe borrowers, and the creditor will then end up having an
adverse selection of borrowers. The situation could be improved through screening and
signaling (Bolton et al., 2005). The creditor can research and screen the borrowers to
determine their eligibility for the credit and also be able to charge a concomitant interest
rate. The borrowers can also signal their borrower types to the creditor by presenting
researched project proposals and records of previous debt contracts and project
performances. This increases the information flow and reduces the adverse selection at
signing of the debt contract. But the borrower may after the contract involve in risky
actions that will be detrimental to the interest of the creditor, and this constitutes moral
hazard. To minimize this problem, the creditor must monitor the actions of the borrower
from the signing of the debt contract to the maturity of the debt. It means, without a debt-
intermediation institution, every creditor should research and screen every loan applicant
and monitor every one of his borrowers. This will entail prohibitive transaction costs for the
economic agents (Ruhle, 1994). Thus, banks emerge to minimize these transaction costs by
intermediating between the creditors and borrowers and acting as both brokers and dealers
in credits. Their total transaction costs of researching, screening and monitoring is lower
- 73 -
than the sum of the transaction costs borne by all individual agents engaged in debt
contracts without an intermediation institution. However, the concentration of credit
functions in the banks can cause them to grow into monopolist institutions that compromise
the efficient allocation of the credit quantities. That is, the debt transaction costs may rise
and the safe projects will be insufficiently financed resulting in some credit rationing and
some adverse selection. Safe projects could also be credit rationed due to supply constraint
– the lending capacity of the banks cannot meet the fund requirements of the safe projects.
The emergence of the banks also does not eliminate the moral hazard. If the banks do not
perform proper monitoring, the bank borrowers may act contrary to the banks’ interest.
Similarly, the bankers may act contrarily to the interest of the depositors. In both cases,
moral hazard occurs. The central bank can, for the welfare of the depositors and the
economy, monitor the bankers, while the bankers must improve their monitoring system
and build incentives that deter the bank borrowers from acting injuriously to the banks.
Let us imagine an economy without debt intermediation. An agent with a financing need of
say 100 € has to search for another agent who has an excess fund and she is willing to lend
out. The first problem for the deficit agent to resolve is to find a surplus agent. This is
termed searching; the deficit agent should search for surplus agents who can satisfy his
financing needs. Similarly, investors (surplus) agents search for potential entrepreneurs,
who can meet the investment requirements of the investors. This search process is
duplicating in effort and time consuming. Duplicating in the sense that, each entrepreneur
has to search for all the potential investors, and vice versa. It is costly and time consuming,
particularly if one investor cannot satisfy the total funds required by an entrepreneur. As the
entrepreneur and the investor are brought together, the searching problem is resolved, and
the second problem appears. That is, the entrepreneur has to be able to secure the funds
from the investor. He has to convince the investor that the project is worth investing in; the
flows from the project can sufficiently pay the principal plus the interest. Meanwhile, the
investor has to examine this claim made by the entrepreneur. This involves high costs for
both the entrepreneur and the investor. The information about the project is asymmetrically
distributed, with the entrepreneur having more information than the investor. This gives rise
to the final problem, which regards how the two parties resolve the asymmetry of
information. it could be that the investor compensates her lack of information about the
- 74 -
project by making her funds highly restrictive and then ends up with only bad entrepreneurs
(adverse selection), or the entrepreneur exploits the lack of information on the part of the
investor and so behaves injuriously to the interest of the investor.
Researching, selecting and monitoring are required to ameliorate the transaction costs and
the consequences of information asymmetry. Because institutions of debt – intermediation
exist to offer researching, selection and monitoring services, which are in turn required due
to the presence of information asymmetry and that the transaction costs of barter trade are
prohibitively costly for both investors and entrepreneurs, then intermediation has to exist.
This existential theory of banking industry asserts that banks are delegated institutions by
investors. After the delegation, the banks research potential entrepreneurs select the
creditworthy ones to minimize the adverse selection, and monitor them to minimize the
moral hazard problem. Thus, banks emerge as middlemen between deficit and surplus
agents in the economy. The theory predicts that any time there is a better arrangement than
banks to ameliorate transaction costs and information asymmetry, the banks will cease to
exist, and as banks engage in the process of providing the three services, they generate
information, employ and train labour, mobilize savings and diversify risk – and this process
is growth promoting.
3.2.2 Evolutionary Theory of Debt – Intermediation
At the early stages of development, financial firms (banks) are not different from non-
financial firms. Banks emerge endogenously developing specific skills for managing risk
and uncertainty. Winkler (1998) surveys the balance sheets of a sample of firms from 1840
to 1989; he shows that “the ratio of equity to total assets was very high at the beginning of
financial system development; and that the assets and liability sides of both financial and
non-financial firms were not significantly different”. Increases in banks’ total assets were
funded by retained earnings. As more equity capital is built via retained profits, the would-
be financial firms begin to lend. The business of lending then caused them to learn the
skills of screening, selecting and monitoring. “They start mobilizing deposits and
accumulating increasingly diversified loan portfolios as their quality skills and capabilities
of screening, selection and monitoring improved” – Winkler (1998). As the process goes
over time the would-be firms then develop into full-fledge financial firms. But this
- 75 -
phenomenon has ceased to reappear. Toady, we do not witness other non-financial firms
being converted to financial firms because they have ample retained earnings and start
lending out. Winkler did not refute the information economics theory of debt
intermediation, he however argues it precedes the information economics theory. He could
have misread the empirical evidence of high equity ratio. High equity ratio on the banks’
balance sheets could be due to a fact that banks were each operating as a currency board,
holding at least 100% reserve requirements against all their issuances. There was no lender
of last resort, and deposits were not guaranteed by a third party. This could force them to
hold high equity capital to back up the deposits and loans they issued.
3.3 Empirical Evidence
Finance-Growth nexus is a long standing debate in both theoretical and empirical literature.
The debate has not shown any sign of a convergent phenomenon, rather it has grown more
diverse and disturbing. Hamilton (1781)8 first introduced the debate by asserting that
“banks are the engine of growth”. Bagehot (1873)9 and Schumpeter (1934)10 had no
problem agreeing with him. The assertion did not run down well with Adams (1819)11, who
came up with an opposing view claiming that “banks (in fact) harm the morality,
tranquillity, and even the wealth of nations”. His view was later revived by Robinson
(1952)12 and Lucas (1988)13 and other post-Keynesian economists, who abstracted their
models from all monetary matters; and banking was viewed of as monetary matters. The
debate subsequently led to four different versions of finance-growth nexus, namely
neutrality of finance, finance retards growth, finance follows growth, and finance causes
growth (Graf, 2001).
8 A quote from Levine, el al. (2000) “ Financial Intermediation and growth: Causality and Cause”, Journal of Monetary Economics 46, 2000, 31-77. 9 Ibid., 3 Ibid, 4 Ibid, 5 Ibid, and 6 Ibid.. .
- 76 -
3.3.1 Neutrality of Finance
If finance is money, and money is a veil, it will not affect the real quantities, such as labour,
capital, consumption and per capita growth rates. This is a most quoted statement of those
who see finance as a neutral factor in real economic development. They measure financial
development as monetary aggregates (liquid assets/GDP, Total Loans /GDP); thus,
increases in banking operations lead to increases in money supply resulting in inflationary
prices. In the long run, financial developments are only inflationary. Fama(1980) applied
the Modigliani-Miller(MM) theorem of irrelevance of pure financing decision to banking
industry. Banks issue deposits and use the proceeds to purchase securities. That is, they
purchase securities (loans) from individuals and firms and then offer them as portfolio
holdings (deposits) to other individuals and firms; Fama (1980) argues that this portfolio
management activity of banks under strong MM theorem is irrelevant to prices and
economic activities. He added; however without proving, that the role of a competitive
banking sector in a general equilibrium is passive. Johnson (1986)14 went the same line of
argument, he asserted that “ a competitive banking system would be under constant
incentive to expand the nominal money supply and thereby initiating price inflation”. The
question to ask is, is finance equivalent to money? Or in other words, is credit equal to
money? The arguments in the favour of the neutrality of intermediation imply that finance
is money; but inside money. They divide money supply into two, inside and outside money.
Money created by central bank is outside money and money created by commercial banks
is inside money. Gurley and Shaw (1960) first invented these terms. They term loans issued
to government as outside money, and only the central bank is empowered to create the
outside money (Bossone and Sarr, 2002). Then, total money supply in an economy should
be the sum of outside and inside money. Nay, this has never been used as total money
supply. Debt-intermediation growth is not synonymous to money supply growth; otherwise,
Japan and Germany with high debt-intermediation shares in the national output should have
experienced inflationary situations over time. To the contrary, these countries have had
stable prices or falling prices as was the case in Japan. Among the tiger economies in Asia,
such as Singapore, South Korea and Malaysia, debt-intermediation has grown substantially
and its share in the output has increased tremendously. But these intermediation 14 A quote from Fama (1980).
- 77 -
developments have no telling effects on the inflation. Thus, if money is neutral and
inflationary, then it is the outside money created by the central banking system not the
inside money created by the commercial banking sector. Banks do not create the inside
money because they want to stimulate the aggregate demand, rather the creation of inside
money is driven by profitability of business projects. This distinction between outside and
inside money can be understood also as primary and secondary credits of Geoffrey
Gardiner15. He made not difference between money and credits; ‘primary credit is newly
created credit and secondary credit is the loans made through the use of assignable debts’.
He argued that ‘ the level of economic activity is determined by three factors, the amount of
new credit created; the speed with which newly created credit circulates, by being spent or
lent; and the rate at which credit is destroyed by the repayment of debt’. In this way, he
sees the trade cycle to be essentially a credit cycle since the credit cannot be expanded
unlimitedly, a boom occurs when the “prudential limit on the creation of new credit is
reached, at that point savers are encouraged to spend so that workers can earn the money
they need to make their desired purchases, otherwise their savings should be allowed to
diminish through inflation”16. As savers spend their savings they monetize the excessive
build up of debt into means of final payments. Thus, if money is credit and credit cycle
constitutes trade cycle, money cannot be viewed neutral in the process of economic growth
and development.
3.3.2 Debt-Intermediation Retards Growth
A close view on neutrality of debt-intermediation gives us an implication that
intermediation retards growth. If neutrality holds, then intermediation will result in
inflationary phenomena and consequently high uncertainty in the economy. High
uncertainty caused by increased inflation increases investment costs causing many projects
to shut down and go out of business due to eroded profits. This will definitely retards
growth. In addition, the proponents of this view argue that the higher the returns debt-
intermediation derives from increasingly better resource allocation depresses saving rates
such that overall growth rates slow down with enhanced debt-intermediation development 15 Credit and state theories of Money, The Contributions of A. Mitchell Innes, edited by L. Randall Wray, University of Missouri , Kansas City, USA, Edward Elgar Publishing Ltd 2004, pp. 157 – 169. 16Ibid pp. 10 – 12.
- 78 -
(Levine et al., 2000). This argument is off the road, efficient intermediation in fact
enhances growth. The intermediations are expected to allocate the society’s scarce resource
(savings) efficiently and competitively such that the loan rate is held at exactly the level
required to equate total loan revenue and total deposit interest costs (Broaddus, 1973),
where the transaction costs are minimum. This activity enhances the saving rates; as
resource allocation becomes more efficient and competitive, banks will be under pressure
to compensate their depositors more handsomely lest they lose them to other debt-
intermediations. Therefore, the argument that high returns from better resource allocation
by banks depress saving rates has no solid ground; because it is only under competition that
banks will undertake better resource allocation and the same competition will force them to
compensate the depositors with higher returns. Winkler (1998) shows that it is only via
efficient allocation of savings that banks can promote growth. For Galetovic (1994)
“intermediation tends to foster the accumulation of physical capital, and doing so it retards
the long run growth and reduces welfare because diminishing returns will eventually set
in”. In other words, the debt-intermediation retards the economic growth through increased
inflation and capital accumulation.
3.3.3 Finance Follows growth
This view has no empirical support. There is evidence of contemporaneous correlation,
which, of course, does not indicate causation. In a response to the question “does financial
activity cause economic growth?” Graf (2001) finds that “ causation runs from finance to
real development with no evidence at all for reverse causation”. That is, economic growth
is not exogenous to the debt-intermediation. However, the proponents of the view that
finance follows growth hold that evolution of intermediation is part and parcel of the
evolution and competition that are accelerating in the markets they serve (Allen et al.,
1993). This implies that debt intermediation is endogenous to the nature of growth process
of the economy. In other words, the level of the intermediation development depends on the
level of economic development. Finance may not lag behind and follows economic growth
if it grows faster than income. Galetovic (1994) finds that “financial systems grow much
faster than income”. This does not convince the opponents, who see the findings of
causations somewhat misleading. That is, most causation evidence was found in already
- 79 -
developed countries with little findings for developing economies. Thus, if finance does not
follow growth but rather leads it, then the causation evidence from developed countries
should be replicable in the developing economies. This has led Leahy et al. (2001) to
conclude that “banks are important at the early stages of the development.” In their
research, they find that adding more recent observations shows that the importance of banks
diminishes.” As the economy develops the role of banks vanishes; they initiate and foster
the economic growth, but their role diminishes as income levels rise and the economy
becomes more efficient and better arrangements of financial intermediation emerge. This
somewhat runs contrary to the empirical evidence that show some degree of causation
running from finance to real sector development in developed economies such as USA,
Germany and Japan with no evidence yet from developing economies.
3.3.4 Finance Causes Growth
Bencivenga and Smith (1991) assert that “the introduction of intermediation in an economy
shifts the composition of savings towards capital, causing intermediation to be growth
promoting”. King and Levine (1993) explore the Schumpeter’s statement that “banker
authorizes people in the name of society as it were to innovate”. This was later augmented
by Taylor (1977)17, who said “the power to allocate capital is one of the most significant
powers in any economic system. It is the power to determine which enterprise will prosper
and which will not”. King and Levine using various measures of financial development in
12 regression equations find that all the indicators of intermediation development are
strongly associated with real per capita GDP growth, the rate of physical capital
accumulation and improvements in the efficiency with which economies employ physical
capital. They hold that association remains economically important and statistically
significant even after controlling for trade, fiscal and monetary policies; and that several
intermediation indicators are good predictors of subsequent growth. They also show that
commercial banks allocate resources better than the central banks, and this is due to the risk
sharing and information services provided by commercial banks. However, the paper’s
assertion that financial indicators can predict economic growth, does not tantamount to the 17 A quote from Ruhle, Ilonka; 1994, “Why Banks? Microeconomic Foundations of Financial Intermediaries”, Franfurt University Dissertation, Peter Lang GHBH, Europaischer Verlag der Wissenschaften, Frankfurt am Main 1997
- 80 -
conclusion that finance causes growth; it may be that finance is only a leading indicator. In
explaining the causality and evidence whether finance is an engine of growth, Jayaratne and
Strahan (1996) observe that rates of real per capita growth in income and output increased
significantly following interstate branch reform in USA. They also note that improvements
in the quality of bank lending not increased volume of bank lending are responsible for
growth changes. Their evidence about bank lending quality relies on bank balance sheet,
the evidence would have been complete and robust had it used the data on bank borrowers,
such as productivity and longevity of the typical bank borrowers. Nevertheless, the finding
was a step forward in ascertaining the causality issue. Rousseau at el. (1998) examine the
intensity of financial intermediation and economic performance in USA, UK, Canada,
Norway, and Sweden over 1870 – 1929, using Vector Error Correction Method, VECM.
They establish a quantitative importance of long – run relationship between intermediation
measures and per capita level of output. They also attempt the causality and conclude that
“output does not feed back directly into intermediation”. Levine and others (2000) later
strengthen the findings of King and others (1993), using a dynamic panel data
econometrics. They show that “exogenous components, such as liquid liabilities and total
assets, of financial intermediation are positively associated with economic growth”. The
exogeneity implies that those components of intermediation, which are exogenous, can
cause economic growth. But the paper does not explain whether the components are weakly
or strongly exogenous; the weak exogeneity may have bidirectional causality. Graf and
others (2001) may also fall into the same trap. They used LISEL method and conclude that
“causation runs from finance to real development, with no evidence for reverse”.
3.4 Conclusions
The survey of theory and evidence shows that there are no conclusive framework of theory
or evidence on the bank-growth interaction. This could be because most theoretical
frameworks hold the Walras’ Law of full information assumption; banks exist because
there exist transaction costs that prevent deficit and surplus agents to communicate and
strike contracts. The assumption of full information practically renders the banks irrelevant
in any theoretical modelling. Also, the formulation of the models and the empirical results
are influenced by the conceptualizations of the developed economies, where the importance
- 81 -
and the role of banks can be assumed away due to the presence of bank-like financial
institutions and high information dissemination level. The situation will be different in
developing economies, where banks are practically the only financial institutions, and they
play a big role in the transmission of monetary policies. Thus, banks are essential in the
early stages of development for both growth financing and the development of other
financial institutions. Banks cannot retard growth unless they refuse competition and
efficiency. Banking development reduces transaction costs and increases competition and
efficiency; a process that works towards near full information. They become irrelevant only
if the full information situation prevails.
- 82 -
4 Theoretical Framework
4.1 Introduction
In this section, we attempt to present some analytical models, without trying to be
exhaustive, that highlight the impact of banking on the national output, price level and
interest. We construct four models that analyze the relationships between the banking
industry and the macroeconomic activities . The first model is the Schumpeter’s circular
flow of creditary production, section 4.2. It discusses how the banks can start the economic
activity. Section 4.3 constructs an econometric analysis for Broaddus (1973)’s competitive
banking industry. It derives some parameters for assessing the competitiveness and
efficiency of the banking industry. Section 4.4 discusses a Keynesian model of small
economy with credit market. It analyzes the interactive impacts between fiscal and
monetary policies and between monetary and creditary policies on the endogenous
macroeconomic variables of output, interest rate, price level and the transaction costs. It
also models a credit constraint and analyzes its impact on the endogenous economic
variables. We define the behavioural relationships based on sections 4.2, 4.3 and 4.4. in
section 4.5. Section 4.6 presents Tsurus’ (2000) model and then attempts a modification of
it, and section 4.7 concludes.
4.2 Schumpeterian model
Schumpeter believes that banks create entrepreneurs who carry out new combinations of
productions that lead to economic growth. He said “the structure of modern industry could
not have been erected without it (credit)”18, and in carrying out new combinations,
“financing’ as a special act is fundamentally necessary in theory as in practice”19. “An
entrepreneur can only become an entrepreneur by previously becoming a debtor”20 to
banks. By not only transferring the existing purchasing power to the fund users but also
creating new purchasing power for financing production, banks act as starters of economic
18 Schumpeter, J. A., 1989(English Translation) “Economic Development: An Inquiry into Profits, Capital, Credit, Interest and the Business Cycles”, Transaction Publishers, p. 70. 19 Ibid 20ibid, p. 102.
- 83 -
growth. This argument then led to development of circular flow of monetary production.
This monetary circuit systematically links the bank money to the production. Banks start
the circuit; however, according to Schumpeter, the money in the circular flow only
facilitates the circulation of the commodities, for the productive investments to be
undertaken banks should create new purchasing power out of nothing - “out of nothing
even if the credit contract by which the new purchasing power is created is supported by
securities which are themselves not circulating media, which is added to the existing
circulation. And this is the source from which new combinations are often financed, and
from which they would have to be financed always, if results of previous development did
not actually exist at any moment”21. This has come to make distinction between banking
and financial markets, banking creates new and additional liquidity to finance carrying out
of production, while financial markets allocate the existing liquidity from investors to fund
users. If additional purchasing power is not created, no change will be effected in the
economy, because in principle, workers and landlords will not loan the services of labour
and land to the entrepreneur. Credit makes it possible for the entrepreneur to buy the
services of labour and land. Thus, banks originate financing, which is fundamentally
necessary for carrying out production. In the Gambia, banking is almost the sole financial
intermediation that together with government appropriation and self finance channels the
national savings to the domestic investments. Then it will be interesting to know how the
disturbances to the banking lending capacity get transmitted to innovations in the private
sector investments. That is, innovations in the private investments should be fully explained
by lending capacity of the banks, the bank credit and the self finance. But self finance is a
privilege that many entrepreneurs will not have. Thus lending capacity of the banking
system and the bank credit must explain significant variations in the private sector
investments; and if the Schumpeter’s credit that finances new combinations exists, the bank
credit will act as an exogenous variable for the private sector investments. Since from the
traditional models, interest rate is an explanatory variable for the investments, I can write
the Schumpeter’s model as follows:
21Ibid., p. 73.
- 84 -
),,(−
= tttt iLCDCPfPI
where, PI = private sector investment, DCP = Bank credit to the private sector, LC =
lending capacity of the banking system, and i = interest rate.
The role of banks in financing entrepreneurs is crucial especially when examining monetary
production of developing countries like the Gambia, where equities markets are almost
nonexistent. If banks cut their supply lines of credits to the entrepreneurs, the production
activities will not be carried out; because no alternative sources of financing such as
equities markets exist. You may ask about self finance or traditional markets. For the case
of self finance, Schumpeter argues that “the possessor of wealth, even if it were the
greatest combine, must resort to credit if he wishes to carry out a new combination, which
cannot, like an established business, be financed by returns from previous production”22.
The individual can achieve the privilege of self finance through savings, not through
property because it cannot be employable either immediately for carrying out new
combination or in exchange for the necessary goods and services. The latter case is not
possible in the Gambia where, most properties are in terms of real estates and livestock,
which cannot be immediately employed in the production and cannot be exchanged for the
goods and services the entrepreneurs wish to produce. Meanwhile, self financing through
savings may not be practical given the levels of income and the low real deposit rates; with
rising inflation rates and the depreciating currencies, the entrepreneurs may not be able to
buy the productive means through savings. Traditional markets, represented by
moneylenders, mostly finance consumption not production goods and services. Thus, bank
credit is essential for production in a developing country; however, the lending capacity of
the bank may be misdirected towards consumption goods and services rather than to
productive investments, and banks can act like moneylenders rather than professional
bankers, giving credits to a clique of friends and high personalities rather to depersonalized
and well appraised projects.
22 Ibid, p. 69.
- 85 -
Schumpeter’s Bank Credit
What determines the credit demand of the entrepreneur given the profit maximizing
principle? For Schumpeter, an entrepreneur demands credit to buy labour hours and capital.
She blends these two inputs using her entrepreneurial skills and produces the output.
To Schumpeter, banks are the circuit starters in the economy; they switch on the engine of
production in the economy as the diagram below illustrates
Source: Bossone and Sarr (2002), WP/02/178.
Banks start the circuit of creditary production by creating credit out of nothing as
Schumpeter has claimed. For example (see the next page for the circular flow balance
sheets), an entrepreneur C, has no finance of his own to buy capital and labor to start
production, can approach the banker to write acknowledgements (bills) to pay for the labor
- 86 -
Balance Sheets of the Creditary Production
b1
C’s debt
integrity
Banker at time t Entrepreneur C at time t
b1 Purchasing Power
Capital Labor
b1
Goods Deposits
Capital Labor
Goods
E & F A & B
Entrepreneur at time t+1
Capital Labor
Goods
Sales = b1 + b2
b1
b1
Profit = d + ib1= b2
Banker at time t+1
b1
C’s debt
b2
I’s debt
b1
E & F’s deposits
b2
d ib1
- 87 -
and capital. The integrity of the banker is accepted by all the market participants. The
banker writes bills, say b1, at time t for C to purchase capital and labor from A and B and
start the production cycle that will end at time t+1. C gives the bills to A and B in payment
and undertakes the production. C becomes a debtor to the banker, and A and B are the
corresponding creditors to the banker. If A and B do not in turn spend the bills but deposit
them in the bank, then A and B are financing the purchases of C. If they spend the bills and
buy goods from E and F, A and B will drop from the bank’s book entry leaving C as the
debtor and E and F as the corresponding creditors. If C at time t+1 can sell his output to E
and F, he will get hold of the b1 and give them back to the banker to cancel his debt, and
correspondingly the credits of E and F are cancelled. This completes the cycle. The
entrepreneur makes zero profit, he spends b1 to finance the production, and sells his output
for b1, which he gives back to the banker to cancel his debt. Also, the banker makes no
interest income, he issues b1 at time t and gets back the same bills at time t+1. If we assume
profit maximization is the objective of the players in this system, then the banker wants the
future value of the b1 at time t+1 charging an interest rate of i, and the entrepreneur must
sell his output at a value higher than b1 at time t+1 at least to break even. Thus, E and F’s
purchases will not clear C’s output, another buyer, say I, is needed to clear the output
balance. I goes to the banker and signs an acknowledgement, say b2 to purchase the output
balance of C. Now, at time t+1, the total costs and sales of C are (1+i) b1 and b1 + b2
respectively. C’s output is cleared, b1 are spent on the output by E and F, and b2 are spent
by I on the output balance; and the total credits created from the beginning of the
production cycle to the end of the production cycle is equal to the total sales of the
production cycle. If b2>i b1, the entrepreneur pays the interest to the banker and makes
some dividend, d. E and F will drop from the bank’s book entry leaving I as the debtor and
C and the banker as the corresponding creditors to the amount of dividend and interest
income respectively. If b2 = i b1, the entrepreneur pays the interest and makes no dividend,
d = 0. E, F and the entrepreneur will drop from the bank’s book entry leaving I as the
debtor and the banker as the corresponding creditor. If b2 < i b1 , the entrepreneur makes a
loss causing the banker to have a non-performing loan. For I to cancel his debt (principal
and interest) with the bank, he must also collect in the course of his business at least the
future value
- 88 -
of b2 at time t+2, and this requires that the banker creates new bills, say b3, so that at time
t+2, I’s collection is b2 + b3>(1+i) b2 to cancel the debt obligations and make some
dividend. That is, for interest and profit making, there must be outstanding debt with
corresponding credit at the bank at every production cycle. If the entrepreneur continues
making profit, the system will be prosperous. If C’s financial need at every production
cycle remains at b1, he will after some time finance his production from his retained
earnings. But this is no good news for C. It means the demand for C’s output remains
constant, and hence C’s production does not grow over the production cycles. If on the
other hand, C improves his technology so that his financial need at every production gets
less and less than b1, then the payments to A and B ( the capital and labor owners) will
become less than b1 at every production cycle. Subsequently, A and B will buy less than b1
from E and F, who in turn will buy less than b1 from C; thus, the demand for C’s output
will be lower than before. If b2 are not sufficient enough to clear the output balance at a
profitable level, C must accept price deflation. Similarly, if A and B save some of b1, C’s
output will be cleared only at deflating prices, if b2 are not sufficient enough to clear the
output balance at a profitable level. The prosperity of the system will depend on the bank’s
credit creation at both the beginning and the end of the production cycle. This depends on
the willingness of the entrepreneur to undertake the production, which in turn depends on
the willingness of the third parties to purchase the output at profitable prices. The
willingness of the players in the system must be realized first, then the bank injects credit
that subsequently creates waves of responses in the entire system. Some of these bills
become money as they become standard and transferable throughout the system. Non-bank
financial inter-mediators, such as stock markets and mutual funds, all deal in already
created bills of the banks. The central bank does the same operation the commercial banks
do but in the name of the government and with restricted customers that are the government
itself and the commercial banks. The government can levy taxes on the citizens and the
commercial banks (via the central bank by ways of regulations) and then deposit these tax
claims in the central bank. The central bank can, against the tax claims, issue credits to the
government. It can also create bills for the government to undertake production. If the
government can collect the future value of the bills at the end of the production cycle, it
will be able to cancel its debt with the central bank and the economy will expand;
- 89 -
otherwise, the value of the bills will fall by the value of the unsold output of the
government production, and the economy will contract. Also, if the commercial banks are
faced with immediate payment of their bills in cash and have no sufficient cash, which they
get through dealing with the central bank, the central bank can issue them cash on credit.
Thus, the ability of the banks to create bills also depends, besides the willingness of the
market players, on the willingness of the central bank to issue cash upon demand. This
concludes the discussion of Schumpeter’s Circular of Creditary Production. We discuss
below in some mathematical terms how the banks’ credits relate to the output and
production.
Assume the technology of production is
labourLcapitalKoutputYwhere
LKfY
:,:,:
)1(),(=
The revenue of the entrepreneur will be PY ( P is the price). The profit is the difference
between the revenue and the costs. The two inputs L and K are paid wage (w) and rental (r)
respectively. The only total cost is wL + rK; and the credit (C) from the bank is exactly
enough to cover this total cost. The entrepreneur has no previous production; and has to
borrow the required funds to cover the total cost from the bank. Thus, this can be
formulated as
)2(rKwLC +=
When Y is produced and sold the production process ends, the entrepreneur pays back the
credit plus interest of one period, takes her profit and winds up the business. This profit is
actually the residual accrued to the owner-entrepreneur of the production. Thus, the
objective of the entrepreneur is to maximize this residual, which we denote it here as
dividend (D). The total cost at end of the period is C + iC, i is the bank’s lending rate; and
this is (1+i)C. The end-of-period costs are
[ ] )3()1( rKwLiTC ++=
The residual D is then
- 90 -
[ ] )4()1(),(),( rKwLiKLpfTCKLpfD ++−=−=
Assuming 1=p (normalized), the entrepreneur maximizes D with respect to the input
values L and K.
)5()1(
,)1(
)1(,)1(
,0/,0/
i
frrif
i
fwwif
getwe
dKdDanddLdDSetting
KK
LL
+=+=
+=+=
==
Substituting the values of w and r in equations (5) into the equation (2), we get the
following:
)6()1()1( i
Kf
i
LfC KL
++
+=
That is, the demand for credit is positively related to marginal productivities of the inputs
and negatively related to the costs of credit (i). This (i) is broadly defined here as all the net
costs the entrepreneur incurs in obtaining the bank credit. If we replace the LLf and KKf
with shares of inputs, assuming α as the portion of national output Y allocated to the
capital; then, (1 - α )Y = LLf , and α Y = KKf . This gives us,
)8(),(
)7()1()1()1(
)1(
−+=
+=
++
+−=
iYCC
i
Y
i
Y
i
YC
αα
Increase in the national output increases the demand for the bank’s credit, while increasing
costs associated with obtaining bank credit depresses the demand for bank credit. Let us
define this cost more formally by expanding the Broaddus (1973)’s model.
- 91 -
4.3 Broaddus’s Model
Broaddus (1973) proposes that banking industry should be opened up for competition to
realize efficient allocation of the scarce resources of the nation. In a competitive banking
industry the loan interest rate is held at exactly the level required to equate total revenue
and total deposit interest costs. That is,
( )9gDRLR =
Where, g = Drr−1
1
Drr = required reserve ratio.
LR = loan interest rate.
DR = deposit interest rate.
Equation (9) is an identical relationship, which we believe does not hold in the real world.
We believe that bank’s lending rate does contain more information than just the deposit
rate, it includes all the costs associated with obtaining the credit. The identity relationship
in equation (9) can be modified to be more representative of the reality in the following
way:
( )10ttt DRLR εβα ++=
This econometric representation tells us that the lending costs of the banks could be more
than the equilibrium value of Broaddus’ model. With β as an estimator of g, and ε
counting for random factors whose expected value is zero, then α can assume values as
follows:
0,0,0 fp ααα and= when the expected lending interest rate is,
( )11)()( tt DRELRE βα +=
- 92 -
Case 1: 0pα , this means the equilibrium deposit costs are higher than the equilibrium
lending costs. That is, banks pay their depositors more than what they reap from the
borrowers. Is this possible? That the real deposit costs are more than the real lending costs.
It also means banks subsidize their borrowers. A monetary policy that suppresses lending
costs to achieve a positive shift of the aggregate demand, while attracting savings by
promising high returns on them may lead to a situation where real lending rate is lower than
the real deposit rate.
Case2: 0=α , this is the case of Broaddus’ model. It represents the perfect equilibrium
situation, where given the high competitiveness of the banking industry the loan interest
rate is held at exactly the level required to equate the total revenue and the total deposit
interest costs. This is the benchmark case for evaluating other cases. Any other case
different from this case is considered a disequilibrium condition, like case1.
Case3: 0fα , this is the other disequilibrium case. Here banks reap a monopoly gain; their
lending costs far exceed the deposit costs. This is the situation we expect to get in this
research; that the banking industry in the Gambia is inefficient, that the costs associated
with obtaining loans from the banks far exceed the costs associated with maintaining the
deposits, that deposit returns are suppressed and as a result less savings are mobilized than
otherwise and in turn resulting in low lending capacity and low loan supply.
Case4: From equation (10), we can derive also an estimate of the required reserve ratio,
Drr . That is, given the lending interest rate and the deposit interest rate, what is the
required reserve ratio? If there is a one-to-one relation between the lending interest rate and
the deposit interest rate, the estimated required reserve ratio, which I call here the empirical
reserve ratio, will be zero. This is the equilibrium point, where the banking efficiency and
competitiveness will hold the prices ( the lending interest rates) of the banking output and
the costs ( the deposit interest rates) of the bank inputs together to co-move in a one-to-one
relation. Banks will not operate at this point because there is a regulatory required reserve
ratio, which is never zero. Nevertheless, they will still operate efficiently and competitively
- 93 -
if the empirical reserve ratio is held equal to the regulatory reserve ratio. This is the
equilibrium point the research will investigate. If the empirical reserve ratio is higher than
the regulatory reserve ratio, the banking industry will be operating at abnormal profit rate
and it will be uncompetitive and inefficient; and the magnitude of the deviation of the
empirical reserve ratio from the regulatory reserve ratio tells us the degree of
uncompetitiveness and inefficiency. This deviation measures the transaction costs in the
bank credit market, zero deviation implies then zero transaction costs in the credit market,
and that results in perfect bank credit market, and this motivates us to measure the
transaction costs as the spread between the lending interest rate and the deposit interest rate.
I combine the implicit outcomes of above two models, Schumpeter’s credit and the
modified Broaddus’s model with Keynesian output, money market and aggregate
production function in the next sections to find out how banks fit into the process of
determining the national income, interest rate, and price level.
- 94 -
4.4 Keynesian Model with Banking Credit
In continuation of the theme developed in the opening section of this chapter, we incorporate
Schumpeter’s credit in a Keynesian model of fiscal and monetary policy analyses. That is, the
Schumpeter’s circular flow of production requires the existence of credit market in addition to
goods market, money market and aggregate production function. The credit market can be
perfect or imperfect, and only bankers are the financial intermediaries in the imperfect credit
market. While, in the perfect credit market, financial intermediation does not exist, because
every surplus agent can costlessly enter into a credit-debt contract with every deficit agent; the
transaction cost is zero. Section 4.4.1 and 4.4.2 introduce the credit market in the Keynesian
IS/LM framework . Section 4.4.1 assumes that in a perfect credit market only the demand for
credit matters and the bank intermediation is irrelevant. Section 4.4.2 introduces credit market
imperfection that brings about transaction costs and consequently the relevance of the bank
intermediation. In this section, we find that bank credit promotes output growth through
investment and this process works also to reduce the transaction costs. The puzzling result is
that as banks develop one reason for their existence in section 3.2. is minimized. Banks exist
because transaction costs exist between the deficit agents and the surplus agents; as banks
develop and expand through increased credit and involvement in investment promotions they
reduce the gap between the two agents by minimizing the transaction costs. This implies that as
the economic development climbs high stages the banking role becomes minimized. Section
4.4.2 also treats credit shocks exogenously that can be enhanced or retarded by a central bank
policy through monetary expansion or contraction respectively. In this section, we treat private
investment as a function of both nominal interest rate and the bank credit. Thus, the impact of a
bank credit is channeled to output through the private investment promotion. Sections 4.4.1 and
4.4.2 examine the interactions between the fiscal policy and the monetary on one hand and
between the monetary policy and the creditary policy on the other hand. In the final section,
section 4.4.3., we introduce the interaction of money and credit in the imperfect credit market
model. We define money here as cash money and demand deposits. Cash money is in the hands
of the public, and demand deposits are in the commercial banks. Cash money also constitutes
part of the bank credit that has been monetized by entrepreneurs. The entrepreneurs monetize
the bank credit when they convert it into a final means of payments. We also treat cash money
and bank credit as substitutes; thus, high cash money holdings reduces the lending capacity of
the banks and consequently the bank credit supply; and reduced bank credit supply will make
the bank credit increasingly costly. If credit is costly, the entrepreneurs will like to hold high
- 95 -
cash balances; hence, high transaction costs associated with obtaining and supplying bank
credit increase the demand for cash money. The other link we introduce is the central bank
money that filters into economy through the bank credit. And the banks convert into credit the
part of central bank advances that have been monetized and deposited back into the banks. In
this process, the central bank money acts as an explanatory variable for the bank credit supply.
Both section 4.4.2 and 4.4.3 assume away the Walras’ Law of full or near full information due
to the presence of imperfect credit market that "arises in a world of uncertainty, a world with
transaction costs and costs of information", (Brunner and Meltzer 1988; Bernanke and Blinder,
1988). The comparative statics analysis in all the following sections are taken from my
supervisor’s textbook, Wohltmann (2000) and manuscripts of the lectures, Money and Credit,
Wohltmann (2005). The model is as follows:
Y = C(+Y) + I(
−i ) + G this is Keynesian aggregate demand, the economic ouput is
distributed among consumption C(Y), Government expenditure G, and the private investment
I(i).
MP
= L(+Y,
−i ) money market, demand for real money balance depends on the income and
the interest rate.
D = D(+Y,
−r ,
−i ) private demand for credit.
Y = income
r = transaction costs. It is the total transaction cost born by both debtors and creditors in
carrying out credit-debt contract.
i = nominal interest rate. In our analysis here, we treat i like nominal discount rate required for
the time value of money.In perfect credit markets, i captures only the time value of money, an
equilibrium discount rate; while, in imperfect credit markets i is assumed to be higher than the
equilibrium discount rate. Both r and i decrease the private demand for credit.
S = S(+Y,
−r ,
+i ), credit supply function. Increases in the nominal discount rate increase the supply
of credit; while increases in transaction costs reduce the supply of credit. Y represents the
economic activity level, the income level and also the lending capacity from the banker’s point
- 96 -
of view. Thus, increases in Y will definitely induce the bankers to increase credit supply.
r = r(−Y,
+P), transaction costs in the credit markets that burden both creditors and debtors.The
costs go down with increasing income and go up with increasing price level
P = price level in the economy. We assume here that this price summarises all the information
available in the economy; a high price could result in an inflationary situation and consequently
high transaction costs; for simplification we use price level instead of price change which is
often used by some researchers (Erhardt, 2002). The difference will be eliminated later when
we take the total differentiations of the log-form of our model. High income ,Y, reduces the
transaction costs. High income increases the value of collaterals, assets and profits and signals
good news about the economic activities; while, low income signals low value of assets.
Agents with high income will encounter less transaction costs than those with low income. The
bankers are more likely to bear higher transaction costs when dealing with low-income agents
than when dealing with high-income agents. This is why we assume that high Y reduces r, and
thus high Y increases the economic agents’ demand for credit bcause it enhances their
creditworthiness via reduced transaction costs and their ability to repay the debt extended to
them. Similarly, high Y induces the creditors to extend the credits due to reduced transaction
costs and improved credit-worthiness of the debtors.
4.4.1 Perfect Credit market
If the credit market is perfect, the total transaction costs will be zero. There will be no friction
cost; bankers will have no place, because economic surplus agents can costlessly meet with the
economic deficit agents and enter into credit and debt contracts. Only the income, Y, and the
nominal discount rate will matter to the contracts; nominal discount rate will be the rate of time
value of money enough to equate today’s consumption with future’s consumption. Under this
situation, the surplus agents will have no problem transferring the surplus to the deficit agents.
The credit will exist only if deficit occurs, the demand for credit determines the credit market;
thus r and S are not required in our perfect credit market model. The banking sector is
represented by the demand for credit, which is purely determined by the entrepreneurs’ demand
for credit, and thus the banking is endogenous when the credit market is perfect. Banking in
this section works only as a filter that filters the deposit money into credit, and since not all
deposit money will be filtered into credit, changes in money may not necessarily equal changes
- 97 -
in credit. This argument will be treated in detail in section 4.4.3 where the money and credit
are linked together via transaction costs and central bank money supply. We construct the
relations in this section as follows:
Y = f(P) following the Keynesian aggregate supply function, given that the labour market is in
disequilibrium,
Nd = Nd−wP
, where, Nd = labour demand. In the short run, when the wage rate, w, is fixed,
the output will be determined by price level.
Our model with perfect credit market is as follows:
Y = C(+Y ) + I(
−i ) + G
MP
= L(+Y ,
−i )
D = D(+Y ,
−i )
Y = f(+P )
In this Model, the credit supply, S, is perfectly elastic and completely determined by the credit
demand, D. After taking the total differentiation of the model, we evaluate the effects of fiscal
and monetary policies, because financial intermediation does not exist in perfect credit market,
we do not have here the effects of credit policy; while demand for credit is assumed to be
always endogenous. The effects of credit policy will be analysed under imperfect credit market.
We incorporate in our analysis the interaction effect between fiscal policy and monetary policy,
and denote that effect as dmdg
under fiscal policy anddgdm
under monetary policy. We further
asssume that this effect is positive if fiscal policy and monetary policy are accommodating (the
two policies take the same direction) to each other, negative if they are opposing (the two
policies take opposite directions) to each other and zero if they are independent.
- 98 -
Fiscal Policy
dy = Cydy + Iidi + dg
dm − dp = Lydy + Lidi
dD = Dydy + Didi
dy = fpdp
Case one: Fiscal policy and Monetary Policy are independent, dmdg
= 0.
dp = 1fp
dy
1fp
dy + Lydy + Lidi = 0
di = 1Ii
(1 − Cy)dy − 1Ii
dg
1fp
dy + Lydy +LiIi
(1 − Cy)dy −LiIi
dg = 0
1fp
+ Ly +LiIi
(1 − Cy) dy =LiIi
dg
We define θ = 1fp
+ Ly +LiIi
(1 − Cy) 0, then
dydg
=LiIiθ 0
Fiscal expansion induces increases in the income. The degree of the income increase is
determined by the terms in θ and the ratio of interest elasticity of money demand to interest
elasticity of investment, the higher is this ratio the higher the degree of income increase
induced by fiscal expansion. From now on, we denote the ratio,LiIi
as ρ.Thus, the impact of
fiscal expansion on the output is positive but less one,
dydg
= ρθ 1 ρ θ
- 99 -
dpdg
= ρfpθ 0 .
Fiscal expansion increases the price level. The higher the ratio ρ the higher the degree of price
increase induced by the fiscal expansion.
didg
= 1Ii
(1 − Cy) ρθ − 1 0
The term (1 − Cy) ρθ 1. This holds since (1 − Cy)ρ θ , where 0 Cy 1 is assumed.
Thus, fiscal expansion increases the nominal interest rate, i, leading to a crowding out effect.
dDdg
= Dyρθ +
DiIi
(1 − Cy) ρθ − 1 0.
The first term on the right hand side is positive; while the last term on the right hand side is
negative, the terms can be viewed as income effect and interest effect induced by fiscal
expansion respectively. If the income effect of the fiscal expanison is higher than the interest
effect of the fiscal expansion on the private demand for the credit, then the fiscal expansion
increases the demand for credit; otherwise, the fiscal expansion will have negative impact on
the private demand for the credit. We believe, the fiscal expansion will have negative impact
on the demand for credit due to increased nominal interest rate that follows the fiscal
expansion. Thus, the fall of private demand for credit will correspond to crowding out effect on
the private investments, as government deficit displaces the private deficit agents.
Case Two: Fiscal Policy with Accommodating Monetary Policy, dmdg 0, or Opposing
Monetary Policy, dmdg 0 .
1fp
+ Ly +LiIi
(1 − Cy) dy =LiIi
dg + dm
θ = 1fp
+Ly+ρ(1 − Cy) 0
then,
dydg
= 1θ ρ + dm
dg 0 : with dm
dg 0,the impact of the fiscal expansion on the output
becomes amplified. The impact of the fiscal expansion will be curtailed when the monetary
- 100 -
policy is opposing dmdg 0.
didg
= 1Ii
(1 − Cy) ρθ + 1
θdmdg
− 1 0 .
Acommodating monetary policy with fiscal expansion are likely to reduce the interest rate.
Now the term (1 − Cy) ρθ + 1
θdmdg
could be greater than one due to the presence of the
term 1θ
dmdg
. That is, ifρθ + 1
θdmdg
is large, there is a possibility for the fiscal policy with
accommodating monetary policy to induce some fall in the interest rate; otherwise the interest
rate will still rise. More precisely, with accommodating monetary policy, the impact of fiscal
expansion on the interest rate is given by this relationship,
didg 0 dm
dg 1
1−Cy1fp
+ Ly .
With opposing monetary policy, dmdg 0, the fiscal expansion will always result in an
increased interest rate and higher than that prevailing in case one.
didg
= 1Ii
(1 − Cy) ρθ + 1
θdmdg
− 1 0 ( dmdg 0 )
Accommodating monetary policy amplifies the impact of the fiscal expansion on the price
level; while opposing monetary policy curtails the impact of fiscal expansion on the price level.
dpdg
= 1fpθ ρ + dm
dg 0 .
The fiscal expansion with accommodating monetary policy tends to increase the private
demand for bank credit, and it tends to reduce it with opposing monetary policy.
dDdg
=Dyθ ρ + dm
dg+
DiIi
(1 − Cy) 1θ ρ + dm
dg− 1 0.
If the monetary policy accommodates the fiscal expansion to the extent that the interest rate
falls, then the private demand for credit will increase. The private investment will be attracted
by the nominal interest rate fall. Otherwise, if the accommodation of monetary policy fails to
reduce the nominal interest rate, the private demand for credit will fall.
Opposing monetary policy can more than reverse or neutralize the effects of fiscal expansion
on both the output and the price level if and only if, dmdg −ρ , or dm
dg= −ρ respectively.
- 101 -
Thus, it is very important for the policy makers to know the value of the rho, because monetary
tightening during fiscal expansion can cause the price deflation and the output fall. But, what is
known is that ρ is often greater than one ( in particular if |Li | → ∞ ,and |Ii | → 0) and this
minmizes the probability that a monetary opposition can offset or reverse totally the effects of a
fiscal policy; while, as we will find later under monetary policy, small fiscal opposition can
reverse totally the effects of monetary policy.
Monetary Policy
dy = Cydy + Iidi + dg
dm − dp = Lydy + Lidi
dD = Dydy + Didi
dy = fpdp
Then,
dp = 1fp
dy
1fp
dy + Lydy + Lidi = dm
di = 1Ii
(1 − Cy)dy − 1Ii
dg
1fp
dy + Lydy +LiIi
(1 − Cy)dy −LiIi
dg = dm
1fp
+ Ly +LiIi
(1 − Cy) dy =LiIi
dg + dm
θ = 1fp
+Ly+ρ(1 − Cy) 0
Then,
dydm
= 1θ 1 + ρ dg
dm,
dpdm
= 1fpθ 1 + ρ dg
dm
- 102 -
didm
= 1Ii
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm
dDdm
=Dyθ 1 + ρ dg
dm+
DiIi
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm
Case Three: Monetary policy and Fiscal policy are independent,dgdm
= 0 .
Monetary expansion will induce an output increase,dydm
= 1θ 0 assuming that rho does not
go to infinity. In comparison to case one, the impact of fiscal expansion on the output is higher
than that of the monetary expansion, if and only if rho, the ratio of interest elasticity of money
demand to interest elasticity of investment demand, is greater than one. That is, in comparison
to case one,
dydg
= ρθ
dydm
= 1θ ρ 1
Similarly, monetary expansion increases price level,dpdm
= 1fpθ 0 but less than that
generated by the fiscal expansion in case one,dpdg
= ρfpθ , if the condition ρ 1 holds. On the
other hand, the effects of monetary expansion on both the output and the pricel level will be
greater than that of the fiscal expansion if the condition ρ 1 does not hold.
The monetary expansion reduces the interest rate, didm
= 1Ii
(1 − Cy ) 1θ 0 . Unlike the fiscal
expansion, the monetary expansion induces positive results in the credit market; the credit
demand increases following a monetary expansion, dDdm
= Dy 1θ +
DiIi
(1 − Cy ) 1θ 0 . This
is due to the reduced interest rate and increased income effect on the credit market following
the monetary expansion.
Case Four: Monetary expansion with accommodating fiscal policy,dgdm
0 , or opposing
fiscal policy,dgdm
0 .
The impact of monetary expansion with accommodating fiscal policy becomes amplified on
- 103 -
both output and price level,
dydm
= 1θ 1 + ρ dg
dm 0,
dpdm
= 1fpθ 1 + ρ dg
dm 0
In a one-to-one accommodating relation,dgdm
= 1 , the maximum impact on the output
becomesdydm
= 1θ [1 + ρ] . The impact of monetary expansion becomes unclear when it
comes to the ineterest rate and the credit market. The interest rate may rise in the presence of
accommodating fiscal policy; that is when fiscal policy expands following a monetary
expansion, didm
= 1Ii
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm 0
if and only ifdgdm
(1−Cy)
θ−ρ(1−Cy) , ordgdm
(1−Cy)1fp
+Ly.
The interest rate will fall in the presence of accommodating fiscal policy if the condition
dgdm
(1−Cy)
θ−ρ(1−Cy) does not hold. Also, if the fiscal policy contracts during the monetary
expansion, the interest rate will fall,
didm
= 1Ii
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm 0 dg
dm
(1−Cy)θ−ρ(1−Cy)
This fall in the interest rate will further improve the effects of monetary expansion on the credit
market and hence increases the private demand for credit.
On the other hand, with opposing fiscal policy, the positive effects of monetary expansion on
the output and the price level are curtailed. These effects are directly determined,
dydm
= 1θ 1 + ρ dg
dm 0, 1
ρ − dgdm
dpdm
= 1fpθ 1 + ρ dg
dm 0 , 1
ρ − dgdm
An opposing fiscal policy can more than reverse the effects of a monetray expansion in a
one-to-one relation,dgdm
= −1 , given that the rho is greater than one. The private demand for
credit will increase because fiscal contraction during the monetary expansion has further
reduced the interest rate. However, if the fiscal contraction during the monetary expansion has
- 104 -
resulted in the reversal of the effects of the monetary expansion on the output, for example,
dydm
0, then the private demand for credit may fall.
dDdm
=Dyθ 1 + ρ dg
dm+
DiIi
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm 0 .
In the special case wheredgdm
= −1 , our reduced solutions become as follows:
dydm
= 1θ [1 − ρ]
dpdm
= 1fpθ [1 − ρ]
didm
= 1Ii
(1 − Cy ) 1θ [1 − ρ] + 1
dDdm
=Dyθ [1 − ρ]+
DiIi
(1 − Cy ) 1θ [1 − ρ] + 1
Assume that 1 − ρ 0 , i.e. ρ =LiIi 1 then the impacts of the monetary expansion
become ineffective, the output and the price will fall, the interest rate and the credit demand
may fall too. Whereas, if we assume that 1 − ρ = 0 , the monetary expansion will have zero
effects on the output and the price, while its effects on the interest rate and the private demand
for credit will be reduced to the following:
didm
= 1Ii 0
dDdm
=DiIi 0
Thus, the policy makers should determine the interaction of monetary policy and fiscal policy,
particularly if the two policies are opposing to each other. This happens often when the two
policies pursue different objectives at a given time period. To minimize the adverse effects of
the opposing fiscal policy, the monetary authorities should make sure that 1ρ − dg
dm, and the
adverse effects of opposing monetary policy is reduced if the conditiondgdm
(1−Cy)
θ−ρ(1−Cy)
holds for the cases of the interest rate the private demand for credit, or ρ − dmdg
holds for the
- 105 -
cases of the output and the price level.
4.4.2 Imperfect Credit Market
D = D(+Y ,
−r ,
−i )
S = S(+Y ,
−r ,
+i )
r = r(−Y ,
+P )
r is the transaction cost that falls with increased level of income, profit, asset value, collateral
value and development level of the economy. We assume all these factors to be captured by Y.
While, we reduce factors such as information uncertainty, country risk, red tape, bribes, side
charges, travel expenses to and fro finance centers and searching and monitoring costs into P ,
price. Price in economics, is the summary of market information. Stable prices reduce the
transaction costs for both the borrower and the lender; while unstable prices increase the
transaction costs.
If the banker is able to totally transfer all the transaction costs to the borrower, the credit supply
will not be affected by r .r will drop from the credit supply function, S = S(Y, i) , and the
banker’s decision on how much credit to supply will be influnced by only the income and the
nominal interest rate or time value of money. Nevertheless, we will not drop the transaction
cost from the credit supply, we derive the reduced solutions and then in the analysis we extract
the consequences of dropping the transaction cost from the lender’s side.
Under imperfect credit market, we have, in addition to fiscal policy and monetary policy, a
credit policy of the bankers. This is because a fiscal policy that affects transaction costs,
interest rate, price level and income level is bound to influence the decision of the bankers on
how much credit to supply. Likewise, a monetary policy that affects the intrerest rate,
transaction costs, the price level and the income is bound to affect the credit supply decision.
On the other hand, according to Schumpeter (1934) and Cameron et al (1967), a banker that
searches for entrepreneurs, organizes them and supplies credit lines to them, has a capacity to
influence the economic activity, transaction costs and prices. This leads us to analyze the
situation where bank credit becomes exogenous opposite to the view that banks follow the
economic development and cannot initiate it. In analyzing the credit policy we include the
position of the monetary policy that is an essential component of our monetary circular flow in
the opening section of this chapter. That is, monetary policy can be accommodating to the
- 106 -
banks’ credit supply or opposing to it. Thus, in this section, we have three policies, fiscal,
monetary and creditary policies. The interaction between the fiscal policy and the monetary
policy is as before, while we furher assume here that monetary policy interacts with the
creditary policy. We first assume that credit supply, S, is an endogenous variable, depending on
income, transaction costs and interest rate, but that there is, unlike the money market, no
equilibrium on the credit market. Thereafter, we assume that S is exogenous, which directly
influences the private sector investments (see subsection 4.4.3).
Fiscal Policy
Y = C(Y) + I(i) + G
MP
= L(Y, i)
D = D(Y, r, i)
S = S(Y, r, i)
r = r(Y,P)
Y = f(P)
Taking the total differentiation of the model:
dy = Cydy + Iidi + dg
dm − dp = Lydy + Lidi
dD = Dydy + Drdr + Didi
dS = Sydy + Srdr + Sidi
dr = rydy + rpdp
dy = fpdp
Then,
dp = 1fp
dy
- 107 -
1fp
dy + Lydy + Lidi = dm
di = 1Ii
(1 − Cy)dy − 1Ii
dg
1fp
dy + Lydy +LiIi
(1 − Cy)dy −LiIi
dg = dm
1fp
+ Ly +LiIi
(1 − Cy) dy =LiIi
dg + dm
θ = 1fp
+Ly+ρ(1 − Cy) 0
Then, as before
dydg
= 1θ ρ + dm
dg
dpdg
= 1fpθ ρ + dm
dg
didg
= 1Ii
(1 − Cy ) 1θ ρ + dm
dg− 1
and therefore,
drdg
=ryθ ρ + dm
dg+
rpfpθ ρ + dm
dg= 1
θ ρ + dmdg
ry +rpfp
dSdg
=Syθ ρ + dm
dg+ Sr
θ ρ + dmdg
ry +rpfp
+SiIi
(1 − Cy ) 1θ ρ + dm
dg− 1
= 1θ ρ + dm
dgSy + Srry +
Srrpfp
+SiIi
(1 − Cy ) −SiIi
dDdg
=Dyθ ρ + dm
dg+ Dr
θ ρ + dmdg
ry +rpfp
+DiIi
(1 − Cy ) 1θ ρ + dm
dg− 1
= 1θ ρ + dm
dgDy + Drry +
Drrpfp
+DiIi
(1 − Cy ) −DiIi
Case Five: Fiscal policy and Monetary policy are independent, dmdg
= 0
Fiscal expansion increases output and price level as in case one,
- 108 -
dydg
= ρθ 0
dpdg
= ρfpθ 0
It also increases the interest rate
didg
= 1Ii
(1 − Cy ) ρθ − 1 0
Also the fiscal expansion may reduce the transaction costs due to the increased impact the
fiscal expansion has on the output,
drdg
= ρθ ry +
rpfp
If |ry | rpfp
, fiscal expansion will lead to reduced transaction costs. This is possible if the
increase in the output induced by the fiscal policy has resulted in reduced transaction costs
dramatically. That could be a situation where the transaction costs are highly sensitive to the
income changes and less sensitive to the price changes, so that the net result of the fiscal
expansion on the output and the price level will produce a reduced transaction cost. Similarly,
the impact of fiscal expansion on the credit supply is not directly identifiable,
dSdg
= ρθ Sy + Srry +
Srrpfp
+SiIi
(1 − Cy ) −SiIi
The impact depends on the transaction costs and the nominal interest rate. If the credit supply
is highly sensitive to transaction costs and insensitive to the nominal interest rate, then the
increased fiscal policy that induces increased transaction costs will cause the credit supply to
fall. The credit supplied may rise if the credit supply is interest elastic and the transaction costs
fall following the fiscal expansion.
Similarly, the fiscal expansion that increases both interest rate and transaction costs can cause
the credit demand to fall,
dDdg
= ρθ Dy + Drry +
Drrpfp
+DiIi
(1 − Cy ) −DiIi
, Dy 0, dD 0, in isolation
The private demand for credit may rise if the transaction costs fall following the fiscal
expansion, and that has been able to more than offset the interest rate effects on the private
demand for credit.
- 109 -
Case Six: Fiscal expansion with accommodating Monetary Policy, dmdg 0 , or opposing
Monetary Policy, dmdg 0 ,
Acommodating monetary policy will amplify the effects of fiscal expansion on all the
quantities. While, opposing monetary policy will curtail the impacts of the fiscal expansion on
the output and the price level. Opposing monetary policy will also cause the interest rate to go
further up than the isolated impact of fiscal policy. This increased interest rate can cause the
private demand for credit to fall, while the credit supply may rise. If ρ − dmdg
holds, then
monetary tightening will reverse the positive effects of the fiscal expansion on the output and
the price level; while its adverse impacts on the credit market variables will be worsened
further.
Monetary Policy
Y = C(Y) + I(i) + G
MP
= L(Y, i)
D = D(Y, r, i)
S = S(Y, r, i)
r = r(Y,P)
Y = f(P)
Taking the total differentiation of the model:
dy = Cydy + Iidi + dg
dm − dp = Lydy + Lidi
dD = Dydy + Drdr + Didi
dS = Sydy + Srdr + Sidi
dr = rydy + rpdp
- 110 -
dy = fpdp
Then,
dp = 1fp
dy
1fp
dy + Lydy + Lidi = dm
di = 1Ii
(1 − Cy)dy − 1Ii
dg
1fp
dy + Lydy +LiIi
(1 − Cy)dy −LiIi
dg = dm
1fp
+ Ly +LiIi
(1 − Cy) dy =LiIi
dg + dm
θ = 1fp
+Ly+ρ(1 − Cy) 0
Then, as before,
dydm
= 1θ 1 + ρ dg
dm
dpdm
= 1fpθ 1 + ρ dg
dm
didm
= 1Ii
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm
and therefore,
drdm
=ryθ 1 + ρ dg
dm+
rpfpθ 1 + ρ dg
dm= 1
θ 1 + ρ dgdm
ry +rpfp
dSdm
=Syθ 1 + ρ dg
dm+ Sr
θ 1 + ρ dgdm
ry +rpfp
+SiIi
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm
= 1θ 1 + ρ dg
dmSy + Srry +
Srrpfp
+SiIi
(1 − Cy ) −SiIi
dgdm
dDdm
=Dyθ 1 + ρ dg
dm+ Dr
θ 1 + ρ dgdm
ry +rpfp
- 111 -
+DiIi
(1 − Cy ) 1θ 1 + ρ dg
dm− dg
dm
= 1θ 1 + ρ dg
dmDy + Drry +
Drrpfp
+DiIi
(1 − Cy ) −DiIi
dgdm
Case Seven: Monetary Policy and Fiscal Policy are independent,dgdm
= 0.
dydm
= 1θ 0
didm
= 1Iiθ
(1 − Cy ) 0
dpdm
= 1fpθ 0
The impacts of monetary expansion on the output, the nominal interest rate and the price level
do not differ from its impacts on the same variables under case one, where the credit market is
perfect and fiscal and montary policies are independent.
drdm
= 1θ ry +
rpfp
If ry +rpfp
0 , the monetary expansion will reduce the transaction costs; otherwise, the
monetary expansion will increase the transaction costs.
dDdm
= 1θ Dy+Drry +
Drrpfp
+DiIi
(1 − Cy )
The monetary expansion that reduces the transaction costs, and the nominal interest rate, will
increase the private demand for credit. Only the termDrrp
fpis a negative value inside the
bracket, the rest are positive, this makes us believe that credit demand will increase following a
montary expansion. But increased prices that follow monetary expansion can distort the
information content of the price level, and as a result the transaction costs may rise leading to
reduced credit demand. Similarly, monetary expansion may or may not increase credit supply,
dSdm
= 1θ Sy + Srry +
SrrpfP
+SiIi
(1 − Cy )
- 112 -
Inside the brackets above, two terms are positive, Sy and Srry, and the other two terms are
negative,SrrpfP
andSiIi
(1 − Cy ) , the net sum of these four terms determine whether or not
the monetary expansion increases the credit supply. Monetary policy, like the fiscal policy, has
no directly identifiable effects on the credit market variables.
Will the situation change if the fiscal policy is accommodating or opposing? With
accommodating fiscal policy, the positive impacts of monetary expansion are amplified, like
the impacts on the output and price level, and the negative impacts are worsened, like the
impacts on the interest rate; while the ambiguity of its impacts on transaction costs, credit
demand and credit supply will still remain. On the other hand, with opposing fiscal policy,
monetary expansion impacts are curtailed, and if 1ρ − dg
dmholds, the opposing fiscal policy
will reverse the impacts of the monetary expansion.
From the analysis so far, it is clear that both fiscal and monetary policies have no meaningful
effect on the credit market variables. The situation becomes worse when the credit market is
imperfect. The monetary expansion tends to increase the credit demand and decrease the credit
supply; while the fiscal expansion tends to reduce the credit demand and increase credit supply.
This ambiguous results tend to support the Schumpeter’s view that bank’s credit is an
exogenous variable that explains other variables such as private investment and consequently
the economic activity. Thus, a monetary policy or fiscal policy may not induce a bank credit
supply.
Banker’s Credit
Looking at our monetary production circular flow at the opening section of this chapter, we see
that actually it is not how much money the central bank releases in the economy, but how much
of that money has been converted by the banks into credit, that matters to the economy. Thus,
the investments (I ) in the economy is not only reducedly determined by the nominal interest
rate but also by the bank credit, I = I(−i ,
+S) , where S = bank credit. In this section, following
the Schumpeter’s circular flow, we investigate the credit impacts on the output, the price level,
the nominal interest rate, the transaction costs and the credit demand. Here we treat the bank
credit as an exogenous policy variable, its value is determined outside the model. In the next
- 113 -
section, we treat it both as a policy variable, and an endogenous variable the value of which can
be determined within the model. We also assume that dg = 0, and it is the monetary policy that
reacts to the creditary policy, and hence dmds
≠ 0 . The model thus looks as follows:
Y = C(Y) + I(i,+S) + G ( S is now exogenous )
MP
= L(Y, i)
D = D(Y, r, i)
r = r(Y,P)
Y = f(P)
Total differentiation of the model,
dy = Cydy + Iidi + IsdS + dg
dm − dp = Lydy + Lidi
dD = Dydy + Drdr + Didi
dr = rydr + rpdp
dy = fpdp, dp = 1fp
dy
(1 − Cy)dy = Iidi + Isds
di = 1Ii
(1 − Cy )dy − IsIs
ds ( now dg = 0)
1fp
dy + Lydy +LiIi
(1 − Cy )dy −LiIs
Iids = dm
1fp
+ Ly +LiIi
(1 − Cy ) dy =LiIs
Iids + dm
θ = 1fp
+Ly+ρ(1 − Cy ) 0
Then
dydS
= 1θ ρIs + dm
dS
- 114 -
dpdS
= 1fpθ ρIs + dm
dS
and therefore,
drdS
=ryθ ρIs + dm
dS+
rpfpθ ρIs + dm
dS= 1
θ ρIs + dmdS
ry +rpfp
didS
= 1Ii
(1 − Cy ) 1θ ρIs + dm
dS− Is
Ii= 1
Ii(1 − Cy ) 1
θ ρIs + dmdS
− Is
dDdS
=Dyθ ρIs + dm
dS+ Dr
θ ρIs + dmdS
ry +rpfp
+DiIi
(1 − Cy ) 1θ ρIs + dm
dS− Is
= 1θ ρIs + dm
dSDy + Drry +
Drrpfp
+Di(1−Cy )
Ii−
DiIi
Is
Case Eight: Credit Policy and Monetary Policy are independent, dmdS
= 0
dydS
= ρIsθ 0
The credit expansion has a positive impact on the output; and this impact is channeled through
the investments financed by the banks.
dpdS
= ρIsfpθ 0
The price level rises as the banks expand credits. This is a natural phenomenon, the creditors
will be able to give increasingly more credits if those taking the credits to finance their
investments expect to receive higher prices for their output. This point will be clear when we
find that expansionary credit will generally increase demand for credit,
dDdS
= ρIsθ Dy + Drry +
Drrpfp
+Di(1−Cy )
Ii−
DiIi
Is 0, ifDiIi
Is is sufficiently
small.
Expansionary credit policy increases the private demand for credit. If the bankers become
entrepreneurs, or promote entrepreneurs, the demand for bank credit will increase. This will be
due to increased information flow about the bank credit that results from reduced transaction
- 115 -
costs particularly on the part of the borrowers.
drdS
= ρIsθ ry +
rpfp
0 ry +rpfp
0
The credit impact on the transaction costs is indirect, it comes from its impact on the output
and the price level. it also depends on the sensitivity of the transaction costs to the output and
to the price level. Thus, we have to know the magnitudes by which the output and the price
level change given a certain magnitude change in the credit supply and know too the sensitivity
of the transaction costs to both the output and the price level for us to precisely determine the
value of drdS
. We resolve this problem in the next section by making an assumption, and later
test the assumption empirically in chapter six. However, we can see today that a banking
system that eleminates minimum balance requirements, or that sends out letters and
information bulletins to their customers and the general public about credit facilities available
in the banking system creates more demand for bank credit than a banking system that does not
do that. As banks take the initiative to create demand and then take measures to meet that
demand, they, in the process, maximize the information flow and minimize the transaction
costs; as a result, allocative efficiency of the credit increases and the credit market works
torwards perfection, But increased credit supply does not imply falling interest rates. In fact,
the interest rate will rise following the creditary expansion. Thus, banks move along their
supply curve of credit and increase the supply resulting in an increased interest rate.
didS
= IsIi
(1 − Cy ) ρθ − 1 0
Case Nine: Credit Policy and Monetary Polciy are accommodating, dmdS 0 or opposing
dmdS 0
If the monetary authorities respond positively to the banker’s initiative by easing the money
torwards the banker’s credit expansion, dmdS 0 , the impacts of the credit expansion will be
amplified; while opposing monetary policy will curtail the impacts of the credit expansion. It is
analogous to the case of fiscal policy with accommodating or opposing monetary policy. The
opposing monetary policy reduces the effects of creditary expansion on the output and the price
level,
- 116 -
dydS
= 1θ ρIs + dm
dSdmdS 0
dpdS
= 1fpθ ρIs + dm
dSdmdS 0
The opposing monetary policy will reverse the effects of the creditary expansion only if
|ρIs | dmdS
. The interest rate will further go up, and the private demand for credit may fall.
Also the opposing monetary policy reduces the effects of the credit supply on the transaction
costs. If the effects of the creditary expansion on the output are reversed by the opposing
monetary policy, then the transaction costs will go up; the interest rate will still rise, while the
private demand for credit will definitely fall. Unlike the monetary policy, the credit policy will
still have a meaningful impact on the output even if the rho goes to infinity, (ρ → ∞) , its
impact on the output will be reduced to ,
dydS
= Is(1−Cy)
Thus significance of the credit policy depends on the relationship between the credit quantity
and the private investment. If the private investment does significantly reacts to the credit
quantity, then the credit policy will have significant effects on the economic activities. The
next section explores this point futher by invoking a neo-keynesian analogy for the modelling
of credit constraint.
4.4.3 Imperfect Credit Market, Money and Credit are Interlinked
In this section, we introduce changes to our money demand function and the credit supply
function in an attempt to explore the interaction of money and credit in our model. The
transaction costs function and aggregate supply function remain as before, while we maintain
our investment to be a function of both nominal interest rate and bank credit. We analyze the
balance sheets of the economy and construct the policy relation that connect money supply,
base money and the banks’ credit supply. An analogy to the Neo-Keynesian theory is used to
explain the modelling of credit constraint.
- 117 -
The Balance Sheets in the Closed Economy
Following Korteweg and Van Loo (1977, pp. 7 - 9), we present all the balance sheets in our
economy. From the balance sheet identities, we derive the relationship between money and
credit and their relation to the base money.
Government
Deposits with the central bank Scd Advances from the Central Bank Scl
Deposits with the banking system Sbd Advances from the banking system Sbl
Loans from the non-bank public Sp
Net worth Wg
The identiy of this balance sheet is Scd + Sbd = Scl + Sbl + Sp + Wg
Total debt of the Government is Scl + Sbl + Sp , and its total deposits are Scd + Sbd ; the net
debt of the government is thus, Scl + Sbl + Sp − Scd + Sbd = St. Adding the net debt to
the net worth of the government, as it can be seen from the balance sheet, gives us zero:
St + Wg = 0, or St = −Wg . We explain the government behaviour later.
Central Bank
Net government deposits Scb Currency outside the banking system C
Advances to the banking system RL Banking system Reserves R
The balance sheet identity: Scb + RL = C + R
The liabilities of the central bank constitute the base money in the economy. It is the currency
outside the banking system plus the reserves of the banking system. Thus, B = C + R. or
Scb + RL = B
It means the monetary base is used to finance the net government deposits with central bank
and the central bank’s advances to the banking system. We can write this as,
Scb = γB, where γ is the proportion of the base money funding the net government deposits
with the central bank, and
- 118 -
RL = (1 − γ)B, (1 − γ) , is the balance funding the central bank advances to the banking
system.
It is the legal money or the high powered money. It has direct link with the government
finance. If the government revenues and expenditures are balanced, the net change of the
monetary base will be zero. For example, when the government makes purchases and issues a
cheque, if the holder of the cheque deposits it in the bank, the reserve of the bank increases by
the amount of the cheque; and if in return the government levies a tax in the amount of the
cheque previously issued, the payer of the tax will withdraw from the banking system the
equivalent amount of the cheque and the reserve of the bank will subsequently fall by the
amount of the cheque, and net change in the monetary base will be zero. That is, the
government has withdrawn from the economy exactly the amount of money it has issued for
the purchases, what has taken place is that some amount of goods and services has changed
hands from the public to the government. On the other hand, if the government purchases
(issues) are higher than its receipts, there will be a net increase in the monetary base. If the
holder of the cheque (assuming now the cheque issues are higher than the cheque receipts of
the government) chooses to hold cash, the currency holdings will increase by the amount of the
cheque. If he instead deposits it in the bank, the reserve of the bank will increase by the amount
of the cheque; and the banking system will end up with excess reserve unless the central bank
cooperates with the government and increases the reserve requirements of the banking system
by the equivalent amount of the government cheque issue, or the banking system offers the
excess reserves in the money market and buys the central bank securities or the government
securities. Also, if it concides with the situation when the banking system needs reserves to
incraese its loans to the private sector, it can then offer the excess reserve as loans to the
private sector. But due to the problem of information, the banking system may not offer excess
reserves completely to the private sector as the government contniues running deficits, the
banking system will give back the excess reserves to those who cause it, the central bank and
the government.
- 119 -
Banking System
Bank Reserves R Demand Deposits D
Net government deposits Sb Time and saving deposits T
Loans to the private sector L Advances from the central bank RL
The identity: R + Sb + L = D + T + RL
Or , Sb + L = D + T + RL − R : The banking system’s earning assets, which here are loans to
the private sector and loans to the government, are funded by demand deposits, time and
savings deposits and advances from the central bank less the reserves. The earning assets,
denoted here EA, are also the credit supply of the banking system. Thus, the bank credit supply
is EA = Sb + L = D + T + RL − R.
From the balance sheet identity of the central bank, RL − R = C − Scb, substituting this in the
earning asset equation of the banking system we get, EA = D + T + C − Scb .
The bank reserves R consist of required reserves the banks must hold against the demand
deposits and time and savings deposits, and the excess reserves the banks wish to hold in
excess of the required reserves. Thus, R is a proportion of D + T, which can be written as
R = k(D + T) , where k is proportional rate.
Non-Bank Public
Currency C Bank Loans L
Demand deposits D wealth Wp
Time and savings Deposits T
Loans to the Government Sp
capiatl stock K
The identity: C + D + T + Sp + K = L + Wp, the assets of the private sector is funded by loans
from the banks and the own wealth. The private sector holds its assets partly as money (final
means of payment) and partly as investments in capital stock, loans to the governemnt and time
and savings deposits. Money as final means of payment gives us the narrow definition of
money. From the balance sheet of the non-bank public, the only assets that can be offered as
final means of payment are currency and demand deposits; and therefore,
- 120 -
M = C + D . Let C = cM be the proportion of money held as currency; then D = (1 − c)M,
(Korteweg and Van Loo, 1977, p.8).
The non-public sector investments are the time and savings deposits, the loans to the
government and the capital assets.We denote these investments as I, then the balance sheet
identity of the non-public sector can be written as money (M = C + D) plus investments
(I = T + Sp + K) on the asset side of non-bank public, which are identical to loans plus wealth
on the liabilities side of the non-bank public,
M + I = L + Wp,
Investments give a rate of return, while money does not give a rate of return. Loans from the
bank have an interest obligation, and the private sector has to pay that interest obligation plus
the loan principal. The entrepreneur, who takes the bank loans, cannot afford to hold the loans
as money, he will rather place them in investments that will give him a rate of return at most
higher than the loan interest rate. Thus, changes in the bank loans will be channelled to the
investments of the entreprenenurs more likely than to his money holdings. If there is credit
rationing, the entrepreneurs will take that into account when deciding how much investments
to undertake not when deciding how much money to hold; thus, credit rationing will spill over
to the investments not to the money market; we will come to this when discussing the
Neo-Keynesian Analogy. Thus, demand for money will follow the Keynesian money demand
function except that we include among prices a price of the transaction costs r, MP
= L(Y, i, r).
On the other hand, the own wealth can be used to fund any of the assets on the balance sheet of
the non-bank public, they can fund with it the loans to the government, loans to the banking
system as time and savings deposits,or finance the capital acquisition or the need for money. If
the need for money arises, the immediate assets that can be liquidated with the minimum costs
are the time and the savings deposits, T; and similarly, if the non-bank public has excess
money, it can easily transfer the excess into time and savings deposits safely with some rate of
return. This gives us the following relation:
T = tM, where t is the proportion of the money kept as time and savings deposits (Korteweg
and Van Loo, 1977, p.8).
From the monetary base through the banking system and to non-bank public we trace the
money supply process as follows:
B = C + R
- 121 -
substituting the value for R from the banking system, we get,
B = C + k(D + T)
Substituting the values for C = cM, D = (1 − c)M , and T = tM from the non-bank public
balance sheets, we get,
B = cM + k[(1 − c)M + tM] = [c + k(1 − c + t)]M , or Money supply is,
M = [c + k(1 − c + t)]−1B = mB, where m = [c + k(1 − c + t)]−1 1
On the other hand, the credit supply process is determined from the earning assets of the
banking system as follows:
EA = D + C + T − Scb
D + C = M, and therefore,
EA = M + T − Scb
substituting the value for Scb from the central bank balance sheet we get,
EA = M + T − γB
Substituting the value for T from the non-bank public balance sheet, we get,
EA = M + tM − γB
Substituting the value for M from money supply process, we get,
EA = mB + tmB − γB = [(1 + t)m − γ]B = aB, where a = [(1 + t)m − γ].
Let a be determined by the interest rate, i, the income level, Y, and the transaction costs, r. The
reserve requirement also determines a, but we assume here that it is constant and hence dk = 0,
and we can leave it out without changing the outcome of the model. Then,
EA = a(Y, i, r)B or an implicit function of EA = S = S(Y, i, r,B),where S denotes the bank
credit supply.
The money supply process and the credit suppply process produce two different multipliers
(Korteweg and Van Loo 1977), which are m and a respectively. The difference occurs because
some money is kept as time and savings deposits, and also because of the central bank
behaviour of allocating the base money between banking system and the government. If the
central bank behaviour is such that γ = 1 ; in other words, the central bank does not give
advances and discounts to the banking system, and t = 0 , no amount of money is transfered to
- 122 -
time and savings deposits, or bank intermediation does not exist, then,
M = B + EA, the money supply will be equal to sum of base money and the credit supply of the
banking system. Here, we do not hold this view, we assume that γ ≠ 1 and t ≠ 0 and hence
money supply is
M = µγB + µEA , where µ = 1(1+t) 1 , or,
M = µγB + µS, where S = EA = L + Sb
From the balance sheet analysis we have the following relations:
M = D + C
MP
= L(Y, i, r)
S = S(Y, i, r,B)
M = µγB + µS
Neo-Keynesian Analogy
Neo-Keynesian theory assumes away the imaginary Walrasian auctioneer, who sets a vector of
prices and wages and takes the resultant demands and supplies and allows no trading until
equilibrium prices and wages are established. When the auctioneer is removed, the agents can
start trading from the onset, as a result the agents will face constraints on their demands and
supplies, because the intial demands and supplies will not match, and thus the quantities
offered or ordered have to be revised to match the actual available ones. Two steps are taken to
reach a state of" rest"and that is the Dual Decision hypothesis. It states that "first the agents
form their plans according to the price signals, in the second step if they are confronted with
quantity constraints they must revise their plans to take into account the qunatity signals as
well’, Felderer and Homburg (1992). For example, the entrepreneur will first plan his
investment according to the interest rate signals, but when faced in the second step with credit
constraints, he must take that too into account. Thus, the effective investment function of the
entrepreneur will depend on both the interest rate and the credit quantity analogous to the
effective consumption function of the Neo-Keynesian theory when the labour faces constraint
in the labour market. The labourer first forms his consumption plan according to the price
signals, but when he finds out that his labour supply cannot be fully employed, he will consider
- 123 -
that fact into his plan because he cannot consume what he has not sold. That is, his
consumption plan cannot exceed his labour sales, the effective consumption plan is that which
takes into account the actual income earned from the labour sales. Similarly, if the
entrepreneur’s investment cannot exceed the credit supply, then there is a credit constraint, and
he can only invest up to the constraint; and thus he must take that constraint into account.
According to Neo-Keynesian theory, if the constrained labourer reduces his labour supply to
the constraint, because the constraint is known with certainty, the constraint is not manipulable
(for example, maintaining higher labour supply than the constraint increases the chances of
getting employment) and there are costs associated with maintaining higher labour supply than
the constraint, a Dreze-demand emerges. That is, the labourer will stop searching for
employment, and his consumption plan will fall to the actual labour sales. The Dreze-demand
has no signal, the labourer has to submit to the situation. On the contrary, a Clower-demand
emerges when the constraint does not stop the labourer from maintaining higher labour supply
than the contraint, because he does not know the constraint for certainty, he believes to get
employment by being optimistic and maintaining higher labour supply than the constraint and
no costs are associated with maintaining higher supply than the constraint (Felderer and
Homburg, 1992, p.215 -216). We assume our entrepreneur is optimistic, he takes the credit
constraint into account when planning his investment. Thus, in the goods market, the effective
aggregate demand consists of effective consumption, C(Y), effective investment, I(i,S), and
government consumption, G. But, at the same time, he maintains a credit demand higher than
the constraint in the credit market. The notional credit demand of the entrepreneurs will be
higher than the realized credit supply. Nevertheless, the entrepreneurs will maintain their
notional demand because they do not know the credit constraints for certainty, and maintaining
the notional demand gives signals to the bankers to increase the credit supply. They do not
have to observe the credit constraint in the credit market, as a result the credit demand will
depend on the level of the economic activity and the prices, D = D(Y, i, r). The effect of the
credit constraint is reflected only in the effective investment function. Thus, from the
Neo-Keynesian analogy we have,
I = I(i, S)
D = D(Y, i, r)
The model is now complete as follows:
Y = C(Y) + I(−i ,
+S) + G : effective aggregate demand
- 124 -
MP
= L(+Y,
−i ,
+r) : money demand
D = D(+Y,
−i ,
−r) : notional credit demand
S = S(+Y,
+i ,
−r ,
+B) : Credit supply, which is equal to the constraint
r = r(−Y,
+P) : transactions costs
M = µγB + µS : Policy variables (M and B are exogenous, and S could also be viewed as
exogenous)
Y = f(p) : Aggregate output
Taking the total differentiation of the model we get,
dy = Cydy + Iidi + IsdS + dg (1)
dm − dp = Lydy + Lidi + Lrdr (2)
dD = Dydy + Didi + Drdr (3)
dS = Sydy + Sidi + Srdr + Sbdb (4)
dr = rydy + rpdp (5)
dm = µγdb + µds (6)
dy = fpdp (7)
then, from equation (7),
dp = 1fp
dy (8)
and substituting (8) in (5) gives us ,
dr = rydy +rpfp
dy = (ry +rpfp
)dy (9)
or, drdy
= (ry +rpfp
) 0 , from the outset, we assume that income increase should lead to
reduced transaction costs, and therefore as an assumption which is close to the reality, we will
hold from now on that (ry +rpfp
) 0. Thus, policy impacts on the transaction costs are
indirect, the policies must affect either output or price level to an impact on the transaction
- 125 -
costs. di can be derived from equation (1),
di = 1Ii
(1 − Cy )dy − IsIi
ds − 1Ii
dg (10)
Substituting the values for dr and dp in equation (2),we get,
1fp
dy + Lydy + Lidi + Lr ry +rpfp
dy = dm
And substituting (10) in the above equation, we get,
1fp
dy + Lydy + Li1Ii
(1 − Cy )dy − IsIi
ds − 1Ii
dg +Lr ry +rpfp
dy = dm
= 1fp
dy + Lydy + ρ(1 − Cy )dy − ρIsds − ρdg + Lr ry +rpfp
dy = dm
= 1fp
+ Ly + ρ(1 − Cy ) + Lr ry +rpfp
dy = ρIsds + ρdg + dm
And let β = 1fp
+ Ly + ρ(1 − Cy ) + Lr ry +rpfp
0, then,
βdy = ρIsds + ρdg + dm , where ρ =LiIi
And substituting the value for dm from equation (6) in the above relation,
βdy = (ρIs + µ)ds + ρdg + µγdb, or ,
dy = (ρIs+µ)β ds + ρ
β dg + µγβ db (11)
Substituting (11) in (10) we get,
di = 1Ii
(1 − Cy ) (ρIs+µ)β ds + ρ
β dg + µγβ db − Is
Iids − 1
Iidg
= 1Ii
(1−Cy )(ρIs+µ)β − Is ds + 1
Ii
(1−Cy )ρβ − 1 dg +
(1−Cy )µγβIi
db
Let α =(1−Cy )(ρIs+µ)
β − Is 0, and ψ =(1−Cy )ρ
β − 1 0, then
di = αIi
ds + ψIi
dg +(1−Cy )µγ
βIidb (12)
Substituting (9) in (3) we get,
- 126 -
dD = Dydy + Didi + Dr ry +rpfp
dy = Dy + Dr ry +rpfp
dy + Didi
Let x = Dy + Dr ry +rpfp
0, then
dD = xdy + Didi
And substituting (11) and (12) in the above equation we get,
dD = x(ρIs+µ)
β ds + ρβ dg + µγ
β db +DiαIi
ds + ψIi
dg +(1−Cy )µγ
βIidb
= x(ρIs+µ)β +
DiαIi
ds + xρβ +
DiψIi
dg + xµγβ +
Di(1−Cy )µγβIi
db (13)
Substituting (9) in (4) we get,
ds = Sydy + Sidi + Sr ry +rpfp
dy = Sy + Sr ry +rpfp
dy + Sidi
And let ω = Sy + Sr ry +rpfp
0, then,
ds = ωdy + Sidi
Substituting (11) and (12) in the above equation, we get,
ds = ω (ρIs+µ)β ds + ρ
β dg + µgβ db +Si
αIi
ds + ψIi
dg +(1−Cy )µγ
βIidb
= ω(ρIs+µ)β +
SiαIi
ds + ωρβ +
SiψIi
dg + ωµγβ +
(1−Cy )SiµγβIi
db, or,
= 1 − ω(ρIs+µ)β −
SiαIi
ds = ωρβ +
SiψIi
dg + ωµγβ +
(1−Cy )SiµγβIi
db
And let φ = 1 − ω(ρIs+µ)β −
SiαIi
0, then,
ds = 1φ
ωρβ +
SiψIi
dg + 1φ
ωµγβ +
(1−Cy )SiµγβIi
db (14)
Substituting (11) in (7), we get,
dp = 1fp
(ρIs+µ)β ds + ρ
β dg + µγβ db , or
- 127 -
dp = (ρIs+µ)fpβ ds + ρ
fpβ dg + µγfpβ db (15)
And finally substituting (11) in (9) we get,
dr = ry +rpfp
(ρIs+µ)β ds + ρ
β dg + µγβ db , or
dr = 1β ry +
rpfp
(ρIs + µ)ds + ρβ ry +
rpfp
dg + µγβ ry +
rpfp
db (16)
Policy Impacts on the Outputdyds
= ρIs+µβ 0
dydg
= ρβ 0
dydb
= µγβ 0
Increased bank credit boosts the output expansion through the intermediation effect (µ), where
the banks mobilize the savings and reduce the money holdings, and through the allocation of
the savings to the private sector investments (Ii ). If the private sector investments are very
sensitive to the bank credit, or similarly if the credit constraint is very strong (Is 1), then
both liquidity trap (Li = −∞) and investment trap (Ii = 0), i.e. , ρ → ∞ will not totally
neutralize the effects of increased bank credit on the output. In this case the credit effect on the
output will be reduced to,
dyds
= Is1−Cy
If (Is 1), the effect of bank credit on the output is greater than that of the fiscal policy on the
output. This could be due to the fact that the bank credit promotes the development of the
private sector where profitability and efficiency are the driving forces, whereas the fiscal policy
could pursue political priorities. The fiscal polciy has also significant effect on the output even
if either liquidity or investment trap exists ρ → ∞ . In this case, the fiscal policy effect
becomes,
dydg
= 11−Cy
- 128 -
Changes in the monetary base are transmitted to the output through the central bank advances
to the government, which will end up in the hands of the public as money payments for the
supply of goods and services to the government. The public could latter deposit this money as
time and savings deposits increasing the lending capacity of the banking system, or they could
place the money in other investments such as loans to the governement or acquisition of
capital. But if there is liquidity trap or investments trap where the public prefers to hold the
money payments as cash and do not place in the investments, then increases in the monetray
base will not have any direct impact on the output,
dydb
= 0, (ρ = ∞)
Policy Impacts on the Interest Ratedids
= αIi 0
didg
= ψIi 0
didb
=(1−Cy )µγ
βIi 0
If there is a strong credit constraint, then any one percentage change in the constraint will be
associated with more than one percentage change in the private sector investments (Is 1). Is
must be sufficiently large in order for α 0. That is, the entrepreneurs will react very
sensitively to any small relaxation in the constraint, and if Is 1 holds, then α 0 is possible,
and hence a positive change in the constraint will be accepted with increased interest rate,
dids
= αIi 0
On the other hand, if the constraint is weak Is 1 (Is is sufficiently small), or the constraint is
not binding, it does not require the entrepreneurs to adjust the investment plans to suit the
constraint (Is = 0), then, α 0 and hence increased bank credit will be accepted only with
reduced interest rate,
dids
= αIi 0
We can reduce the above two inequalities as follows:
- 129 -
dids
= αIi 0 α 0 (1 − Cy )µ − 1
fp+ Ly + Lr ry +
rpfp
Is 0, thus, α
can be positive if and only if Is is sufficiently small or zero, which means no credit constraint.
Fiscal policy, as in other cases before, increases the interest rate; while increases in the
monetary base reduces the interest rate. If the increase of the monetary base consists of only
advances to the banking systems (γ = 0), the interest rate my not change. Whereas, if the
increase consists of only advances to the government, the interest rate will fall; the fall could be
curtailed if the banking system could attract later those advances as savings from the third
parties (µ 1).
Policy Impacts on the Demand for Bank Credit
dDds
= x(ρIs+µ)β +
DiαIi
0
dDdg
= xρβ +
DiψIi
0
dDdb
= µγβ x +
(1−Cy )DiIi
0
Increased bank credit create demand for the bank credit. With credit constraint, any triggered
relaxation of the constraint will cause the entrepreneurs to put up higher demand signal to
increase the chances of their demands being met. Fiscal policy could crowd out the
entrepreneurs due to the increased interest rate that follows a fiscal expansion. There is
however a possibility for crowding-in effect, wherexρβ
DiψIi
, which means the income
efffect following the fiscal expansion is higher than the interest rate effect on the private
sector’s demand for the bank credit. The increase in the monetary base aslo increases the
demand for bank credit due to the reduced interest rate and the outptut expansion that follow
increased monetary base.
Policy Impacts on the Bank Credit Supply
dsdg
= 1φ
ωρβ +
SiψIi
0
- 130 -
dsdb
= µγφβ ω +
(1−Cy )SiIi
0
Increased government spending could reduce the bank credit supply to the private sector due to
the crowding out effect, where the government pays higher and higher interest rates for the
bank credit. The situation will be worse if there is already a strong credit constraint, the lending
rate will soar up and the credit supply to private sector will drastically drop, because the
government cannot be rationed; this could result in total crowding out of the private sector. A
possibility for crowding in effect on the credit supply of the banking system will only exist if
the credit constraint is very weak. Under strong credit costraint, the monetary policy expansion
will promote credit supply expansion, while under weak credit constraint, monetary policy
expansion may not change the bank credit supply. The banking system may reduce its credit
supply to the private sector following a monetary policy expansion, if the credit constraint is
very weak and the resultant price level increases obscure the credit market information. But the
impact of monetary policy (here monetary base changes) depends very much on the
intermediation (µ) and the proportion of the monetary base that is held as advances to the
government (γ). If the changes in the monetary base are mere advancses and discounts to the
banking banking system (γ = 0) , the bank credit supply may not necessarily change. The
changes in the monetary base will be effective if they are advances to the government, and the
banking system later attract those advances from the third parties as savings and deposits.
Policy Impacts on the Price Leveldpds
= (ρIs+µ)fpβ 0
dpdg
= ρfpβ 0, and
dpdb
= µγfpβ 0
Credit expansion, fiscal expansion and monetary expansion all lead to price level increases. In
the presence of strong credit constraint (Is 1), credit expansion will generate higher price
level increase than the fiscal policy can generate, or the monetary expansion can generate.
Whereas, if the credit constraint is very weak (Is = 0), fiscal expansion will generate higher
price level increase than the credit expansion or the monetray expansion can generate. The
impacts of credit expansion and monetary expansion on the output will coincide, if the credit
constraint is very weak (Is = 0) and the central bank does not give advances and discounts to
- 131 -
the banking system (γ = 1),
dyds
= µβ = dy
db, and
dpds
= µfpβ = dp
db
The results above will never hold if either Is 0 or 0 γ 1
The impact of credit policy is optimal if ds = dD . Generally, credit supply is constrained, the
quantity of credit the banks offer is often below what the entrepreneurs demand. This can force
the entrepreneurs to scale down their investment options leading to a situation of
underinvestment, and hence ds dD; the credit impact on the output will be subsequently
suboptimal. The credit could be also demand constrained that the entrepreneurs have no
additionnal profitable investments, or they have excessive debts that prevent them from taking
all the debts the banks are willing to offer. This situation is rare in developing economies
where the investments options are yet to be fully utilized. It will be often a supply constraint,
low lending capacity and the problems of asymmetry of information that hold the credit supply
below the optimal level of the economy’s demand. Therefore, investments here react to both
interest rate and the quantity of credit, if the investment elasticity of credit is greater than zero.
If the credit constraint is very strong, (Is 1), an increased credit supply may be assoicated
with increased interest rate. If the assumption |Di | |Si | holds and the investment elasticity of
credit is between zero and one, 0 Is 1, increased credit quantity may be followed by a fall
in the interest rate, and this also means the credit constraint is weak. Thus, if credit constraint is
weak, or not binding, banks can increase their credit supply to the private sector only by
reducing the interest rate. In chapter 6, we examine empirically whether these models,
assumptions and hypothetical relationships hold in the case of the Gambia. The derived
hypothetical relations that ensure significant impacts of bank credit on the economic activities
from this section are,
0 Is : Credit constraint exists
0 Is 1 : Weak credit constraint
Is 1 : strong credit constraint
Lr 0
ry 0
rp 0
- 132 -
|Di ||Si |
The relationship between credit and investment will be examined in the context of
Schumpeter’s credit, where bank credit is claimed to be causal for the entrepreneurial
investments. In analyzing the bank’credit market, and if the credit demand and supply
functions are identifed, we will know whether or not the demand curve is flatter than the supply
curve. Also in estimating the function of real money demand, we can examine the relation
between the real money demand and the transactions costs; and the assumptions that increased
real income reduces the transactions costs and increased price level increases the transactions
costs can be tested if an equilibrium relation exists for the linear combination of transactions
costs, real income and the price level. Meanwhile, the total credit of the banks, the total liquid
liabilities and the total assets of the banks will be taken as measures for the roles of banks as
engines of growth. They provide credit for the economy to finance investments and purchases
of goods and services. Their liquid liabilities function as an efficient payment system for the
economy, and their asset accumulation over time promote the capital formation. The ability of
the banks to operate these functions sustainably and efficiently over time should expand the
national output, create and promote investments, increase the capital formation and induce
improvements in the economic efficiency. Thus, the roles of banks will be empirically
examined with respect to the national output, GDP, the domestic investments, the capital stock
and the improvements in the economic efficiency. We present and discuss these behaviorial
relationships in the next subsection.
4.5 The Behaviorial Relationships
The relationships in this research are analyzed using Vector Error Correction Method (VECM)
of Johansen(1995) procedure. The four information criteria of Final Prediction Error, Akaike
Information Criterion, Schwarz Criterion and Hannan and Quinn Criterion are used to choose
the optimal lag length for the VECM. That is, we first run an autoregressive model for each
relationship and compute its optimal lag length of say (q), and then the optimal lag length for
the VECM will be automatically (q-1), see Luetkepohl(1991). The unit root test for each
variable is conducted using the Augmented Dickey-Fuller method; see Gujarati (1995) . The
behaviorial relationship will be presented and discussed if there exists a co-integration relation;
that is, we are concerned here with the long run interactions among the variables. If two
variables are co-integrated, the economic forces will not allow them to wander apart
- 133 -
permanently. These economic forces are thought of as implying long run equilibrium relations.
Thus, there would be only temporary deviations from the long run equilibrium, despite the
tendency for each variable to individually follow a random walk. With cointegration vectors,
we can analyze the responses of the variables to a shock to one variable in the system. This
shock is analyzed in the impulse-responses. With a "cointegrating relation, a shock to one
variable will not affect only that particular variable but also the entire system of the variables"
(Luetkepohl, 1991, p.133). I use the Granger causality results to evaluate the impulse responses
because " an innovation in the variable k has no effect on the other variables if the k variable
does not Granger cause the set of the remaining variables" in the system (Luetkepohl, 1991,
p.45). The data are collected from various sources, and we use annual data throughout the
econometric estimations. The data come from World Development Indicators 2003 /World
Bank, International Financial Statistics 2003/IMF, World Statistics/UN , Balance of Payments
2003/IMF and various publications of the Gambia Central Bank . The sample size is 1964
-2002, and it is applicable to all our econometric estimations. We use E-Viwes 4 to produce the
econometric outputs.
The VAR models, Zit (i = 1,2, . , . , 12) of the research that form the relationships are as
follows:
1. Gross Domestic Product and the Banking
Z1t = (LRYt,LRPDCPt,LRPLLt,LRPCBAt)
where,
LRY = log real per capita output /GDP.
LRPDCP = log real per capita banks’ credit to the private sector.
LRPLL = log real per capita liquid liabilites.
LRPCBA = log real per capita commercial banks’ assets.
2. Private Investment and the Banking
Z2t = (LRPPIt,LRPDCPt,LRPLLt,LRPCBAt).
LRPPI = log real per capita private investments.
3. Gross Domestic Investment and The Banking
Z3t = (LRPGDIt,LRPDCPt,LRPLLt,LRPCBAt).
- 134 -
LRPGDI = log real per capita gross domestic investments.
4. Physical Capital Stcok and the Banking
Z4t = (LRPKt,LRPDCPt,LRPLLt,LRPCBAt).
LRPK = log real per capita physical capital.
5. Economic Efficiency and the Banking
Z5t = (EFt,LRDCPt,LRLLt,LRCBAt).
EF = growth rate of improvements in efficiency. This is here the residuals from aggregate
production function. King and Levine (August,1993) decomposed the sources of economic
growth into two components, physical capital accumulation and everything else. They define
the production technology as follows:
Yt = KtαXt
Y = output
t = time
K = Physical capital accumulation
X = Everything else.
α= capital share
X are other sources of growth, such as human capital, technology and improvements in the use
of factor inputs. They used real per capita values. If y, k and x are growth rates of real per
capita output, real per capita physical capital accumulation and real per capita everything else
respectively, then taking logarithms of equation (1) and differentiating, according to King and
Levine, we get ,y = αk + x. x was defined in their paper as improvements in efficiency,
because it measures the residuals of real per capita GDP growth after accounting for the growth
rate of physical capital accumulation.Thus, if α is known, x could be calculated, and depending
on transformation of GDP and K, X also will be automatically transformed accordingly. For the
calculation of the efficiency, I use the real per capita values and the real per labour values of
equation (1). α of 0.3 is used in all the calculations, and the technology is assumed to be 1.
Thus, if rpef and refl are log real per capita value and log real per labour value of X
respectively, we can calculate them as follows:
Lry = αLRPK + rpef
Lryl = αLRKl + refl
- 135 -
where,Lry = log real per capita output, LRPK = log real per capita physical capital, Lryl = log
real output per labour, LRKl = log real capital per labour
LRDCP = log real banks’ credit to the private sector.
LRLL = log real liquid liabilities.
LRCBA = log real commercial banks’ assets.
6. Economic Activities and the Banking
Z6t = (LRYt,LRPPIt,LRPKt,EFt,LRPDCPt,LRPLLt,LRPCBAt)
From VAR 1 to 5, three banking factors are analyzed with respect to individual economic
variables. In these VARs we expect to have one cointegration relation, which corresponds to
the long run equilibrium relation that connect an individual economic variable with the linear
combination of the banking factors. VAR 6 groups the economic variables together with the
banking factors in an ad hoc ordering manner. That is, the economic variables are ordered first
followed by the banking factors; and we expect to have four cointegrating relations. With four
relations, it means each cointegrating row represents an equilibrium relation between the
respective economic variable with the linear combination of the banking variables. The VAR 1
to 4 and VAR 6 are estimated using log real per capita values, and VAR 5 is estimated using
log real values.
7. Commercial Banks’ Credit Market
Z7t = (LDCPt,LRt,DRt,LLCt,LGDSt,LGDPt)
LDCP = log banks’ credit to the private sector.
LR = lending interest rate.
DR = deposit interest rate.
LLC = log lending capacity of the commercial banks, which is here the total
liabilities of the banks
LGDS = log squared quantity of the banking sector’s claims
on the government.
LGDP = log Gross Domestic Product.
Z8t = (LRMt,LRGDPt,LRt,IDt)
LRM = log real money supply, M1.
- 136 -
LRGDP = log real GDP.
ID = the spread between the lending interest rate and the deposit interest rate.
Z9t = (IDt,LRGDPt,LCPIt)
LCPI = log consumer price index.
VAR 7 represents the supply and demand functions of the commercial banks’ credit. After the
determination of the cointegration rank, we will introduce restrictions on the cointegrating
vectors and the loading coefficients to identify the two functions. VAR 8 is used to analyse the
relationship between real money balance and the transaction costs. We have it in the theory that
the entrepreneurs will demand higher real money balances the higher the transactions costs
associated with bank credit. VAR 9 is employed to examine our assumptions in Chapter 4 that
the higher the income level the lower the transactions costs, and the higher the price level the
higher the transaction cost.
8. Competence of the Central Bank Monetary Policy
Does the central bank sterilize inflows of net foreign assets to achieve its monetary target?
Does it withhold or reduce domestic credit when domestic inflation exceeds world inflation?
For example, squeeze money supply through open market operations, and other means, to keep
in line of its target.The importance of the role of the central bank in the economy depends on
the effectiveness of its monetary policy. The research on banking-growth nexus will be
incomplete unless the role of the central bank is examined. This role is theoretially explained in
chapter 4. We use Cumby and Obstfeld(1981) model in the following VAR and then estimate
the vector error correction method:
Z10t = (LDCt,LNFAt,LCAt,LEXt, INGAPt,LDCGt,LOIPt,LGDt,LGDSt),
LDC = log domestic credit.
LNFA = log net foreign assets.
LCA = log current account balance.
LEX = log exchange rate index.
INGAP = inflation gap between the Gambia and the UK inflation rates
LDCG = log domestic credit to the government.
LOIP = log Dubai crude oil price index.
- 137 -
LGD = log government debt.
LGDS = log government debt squared.
A complete sterilization of net foreign assets implies a negative one-to-one reltaion between
the log domestic credit and the log net foreign assets, Frankel (1994). Thus, if there exists a
long run relation, then making the restrcition that the summation of the respective long run
coefficients of log domestic credit and the log net foreign assets must equal to zero for the
complete sterilization of net foreign assets or total ineffectiveness of the monetary policy. The
variables, DCt,NFAt, and CAt will be transformed as percentages of GDP to avoid
encountering log of negative values and also to take account the effects of the national output
on the these variables. furthermore, our data insufficiency does not allow us to estimate the
above system with all the variables, and we therefore estimate only a bivariate co-integration
between domestic credit and net foreign assets.
9. Schumpeter’s Credit
Z11t = (LRPIt,LRDCPt,LRLCt,LRt)
LRPI = log real private investments
LRDCP = log real banks’ credit to the private sector.
LRLC = log real lending capacity of the commercial banks.
LR = lending interest rate.
Schumpeter has made direct link between the entrepreneur’s activity and the bank credit. We
present this VAR and estimate the VECM to test the link betweeen the private investment and
the bank credit. If banks do anticipate the entrepreneurial activity, there will be feedback effect
from the bank credit to the private investment. While, if banks follow the entrepreneurs, the
private investment will feed back to the bank credit. The lending capacity, controls for the
willingness of the banks to supply the credit, and lending rate captures the price for supplying
the credit.
10. Broaddus’s Efficient Allocation of Resources by Banks
Z12t = (LRt,DRt)
LR = lending interest rate.
DR = deposit interest rate.
The details of this model are presented in section 4.3.
- 138 -
4.6 Tsuru’s(2001) Model of Financial Intermediation Tsuru links the financial intermediation to economic growth using AK (A is the
productivity of capital and K is the physical capital stock) endogenous growth model. His
model has left many unanswered questions. We first present his model and then introduce
the unanswered questions.
He assumes a typical AK model, where the technology of production is
Y = AK
Y = the output
A = productivity of capital
K = capital stock
After his mathematical derivations, the growth rate of capital is given by the following:
δθ −= sAq\
where
=\q growth rate of capital per worker
=A productivity of capital
=s saving rate
=δ depreciation rate
=θ portion of savings used for investments, 10 pp θ
The model argues that banks influence the portion of savings used for investments. That is,
savings are filtered through the banks to become investments. The quality and efficiency of
banks determine how much savings can be processed into investments output. Before
discussing the model, let us construct the model using both AK model and neoclassical
Cobb-Douglas model taking into account the labor growth which is left out in the Tsuru’s
construction:
AK model
)1(),( LKfAKY ==
( ) ( )2,\
KLKsfIK δ−==
L = labor
- 139 -
K\
= change in the capital stock
I = investment
Tsuru asserts that without financial intermediation, KLKsfIK δ−== ),(\ ; while in
the presence of financial intermediation, it is a portion )(θ of ),( LKsf that is channeled to
investments. That is, ),( LKsfθ is the investment, and ),()1( LKsfθ− is absorbed by the
intermediation as information and transaction costs, spread between lending and deposit
rates, commission and transaction fees and operating expenses. Thus,
)3(),(\ KLKsfIK δθ −==
Let r be capital labor ratio. Then we have the following to complete the dynamics of the
AK model:
)4(L
Kq =
Then, the intensive production function is,
)5()()1,( AqqfL
Kf
L
Yy ====
If the labor population grows according to the exponential function of,
)6( ent
L =
where, n = population growth rate, and t = time; and assuming L0 = 1. Then the growth rate
of labor, L, is,
( )7
\
nLL =
Differentiating equation (4) with respect to time we get,
)8( ,\\\
2
\
2
\\
nqqL
Kornq
L
K
L
KL
L
LKq +=−=−=
- 140 -
From equation (3),
)9( )(\
qqsfL
I
L
K δθ −==
From equations (8) and (9): left-sides are equal, then the right-sides are equal; thus,
)10( )()(,)(\\
qnqsfqorqqsfnqq δθδθ +−=−=+
From equation (5): Aqqf =)(
Then,
)11( )(,)(
\\
δθδθ +−=+−= nsAq
qorqnsAqq
This is graphically presented in figure 4.2,
Figure 4.2
That is, as long as )( δθ +nsA f the growth rate of capital will be positive. In fact, banks
can increase the magnitude of this growth rate by filtering more savings into investments
than any time before. They also hinder and reduce this growth rate by withholding high
portion of savings and spending them on the operations and maintenance of the banks.
Since this behavior of the banks are directly linked to the growth rate of capital, which in
turn is equal to the growth rate of per capita output of the economy, then their behavior is
important to the economy, and any fluctuation in their behavior of filtering the savings will
cause fluctuations in the entire economy.
sAθ
)( δ+nq
q\
q
- 141 -
We can also present Tsuru’s model in a neoclassical production function that exhibits a
constant returns to scale, with positive inputs, K>0, L>0, it exhibits positive and
diminishing marginal product with respect to each input, and it fulfils the Inada conditions
that marginal product of K ( L ) approaches infinity as K ( L ) approaches zero and
approaches zero as K ( L ) approaches infinity ( Barro and Sala-I-Martin, 1995). Thus,
unlike the AK model, the neoclassical production function does not have a constant growth
rate of capital per labor. Equation (11) becomes,
)12()()(\
qnqsfq δθ +−=
while )(qf is the intensive form of the production function.
Then growth rate of capital per worker is,
)13()()(
\
δθ +−= nq
qfs
q
q
Figure 4.3
The fraction of resources )1( θ− absorbed by the financial intermediation is indispensable
for the system to operate. But according to Tsuru, this fraction could be set at inefficiently
high levels due to monopoly power, regulation or other reasons. If these quasi-rents
extracted by the financial intermediation are spent on private consumption or inefficient
)(qf qn )( δ+)(qf
)(qsfθ
*
q
q
- 142 -
investments, the loss of resources )1( θ− depresses the economic growth; otherwise the
growth is promoted and enhanced by the presence of financial intermediation.
Tsuru’s explanation failed to relate the presence of financial intermediation to saving rate
and output; is the saving rate the same before and after the presence of financial
intermediation? Or does the output function shifts as the intermediation emerges in the
economy? And how does the fraction absorbed by the intermediation behaves over time?
Without elaborating on these unanswered questions, Tsuru’s introduction of intermediation
makes the economy worse off, no matter how efficient is the financial intermediation. See
the graphs below:
Figure 4.4: An economy without financial intermediation
)(qf )(qf
)(qsf
0qeq
qn )( δ+
q
- 143 -
Figure 4.5 : The economy with financial intermediation
Comparing figure 4.4 and figure 4.5, we can see that the equilibrium capital per worker is
lower in the presence of financial intermediation than that without the financial
intermediation given that )(qf and s do not change as the intermediation is introduced.
Since θ cannot be greater than one, the presence of intermediation makes the economy
worse off; the absorption of savings by the intermediation reduces the investments. The
intermediation can achieve the equilibrium point of figure 4.4 only by channeling all the
savings into investments, )1( =θ . The other way to achieve figure 4.4 equilibrium point is
to consider any absorption by the intermediation to be productive investments; so that what
are actually saved, S, through the banks are all invested, )()()1()( qsfqsfqsfS =−+= θθ ,
which is exactly the same like that prevailing in the absence of intermediation. The model
has posited interesting questions, and the importance of the model lies in elaborating those
questions. This is possible through small modifications of it.
Modifications to the Tsuru’s Model:
In the presence of banks, or a bank-based economy, the national savings are channeled to
investments through two ways, bank and non-bank intermediations. The non-bank
intermediations are direct lending to the entrepreneurs, the funds are channeled directly
from savers to the entrepreneurs. This saving will be relatively small in the presence of
)(qf
)(qsf
)(qsfθ
*q eq 0qq
)(qf
qn )( δ+
- 144 -
banks, it will be high in the absence of banks, because in that case all surplus agents must
invest their funds directly or do no investments. Thus in the presence of banks the
investments in the economy becomes )()()( 21 qfsqfsS θ+= , with )(1 qfs as the
investments through non-bank intermediation and )()(2 qfs θ as the investments through
bank intermediation. The rate of savings channeled via the banks to investments is
dependent on the efficiency of the banks; 0)0(2 =s means the banks are completely
inefficient, and all that is saved through banks never makes to productive investments;
while 1)1(2 =s means the banks are perfectly efficient, all that is saved through the banks is
being channeled to productive investments.
In the case of financial autarky, total savings is )(qsfS = . We assume that these savings
are greater than )(1 qfs , the savings or investments by the non-bank sector in a bank-based
economy. Thus, the total savings in a bank-based economy, )()()( 21 qfsqfsS θ+= , can
be less than, equal to or greater than the total savings in a non-bank-based economy,
)(qsfS = , depending on how efficient are the banks in financing the productive
investments. The graphs below illustrate these cases.
- 145 -
Figure 4.6: Case 1
Case 1: sss f))(( 21 θ+
The graph above shows a case where banks are efficient at both mobilizing and directing
savings to productive investments. This has resulted in a higher equilibrium capital per
worker, , than that prevailing in the non-bank-based economy. The production technology,
)(qf , in both economies is the same; but banks’ intermediation has raised savings rate,
and the mobilized savings by banks are not wasted on the absorption.
)(qf
)(qf
)())(( 21 qfss θ+
)(qsf
q0q2q1q
qn )( δ+
2q
- 146 -
Figure 4.6: Case 2
Case 2: sss p))(( 21 θ+
We have already assumed from the outset that ss p1 when the production technology is the
same in both economies. Therefore, the possibility for the total saving rate in a bank-based
economy to be higher than that of the non-bank-based economy is determined by the level
of banks’ mobilization and efficiency. The case above illustrates a situation where banks
release only small amount of liabilities into investments; that is, the θ is low, or their
mobilization of liabilities, 2s , is low. In both situations, the equilibrium capital per worker
in the bank-based economy will be lower than that of the non-bank-based economy.
qn )( δ+)(qf
)(qf
)(qsf
)())(( 21 qfss θ+
q1q2q 0q
- 147 -
Figure 4.7 : Different production functions
fB = production function in the bank-based economy
fNB = production function in nonbank-based economy
Case 3: the production technologies are different and bank-based economy has higher
production technology, , than non-bank-based economy . Thus, even if the
rates of savings are the same, sss =+ ))(( 21 θ , in both economies, the total per capita
saving in the bank-based economy will be higher than that in the non-bank-based economy
resulting in a higher equilibrium capital per worker in the former than in the latter. This is
possible according to Joseph Schumpeter if the banks finance not only the existing lines of
production but also create new lines of production by selecting and sponsoring
entrepreneurs.
But can this portion , )(2 θs , be allocated to another type of financial intermediation that
can do better than the banks, because it has lower absorption costs, )1( θ− , than the banks?
The natural answer is yes, stock exchange has lower absorption, it is an intermediation
where funds from investors are directly channeled to the users with minimal absorption.
)(qf B
)(qf NB
)(qf
)(qsf NB
qBqNBq2q1q
qn )( δ+
)())(( 21 qfss Bθ+
)(qf B )(qf NB
- 148 -
But do stock exchange mobilize savings or sponsor entrepreneurs? This is where the banks
make difference. Banks create the liquidity in which the stock exchange operates, by either
underwriting stock issues of companies or by sponsoring the stock dealers and stock
buyers. Stock exchange is a complement to the banks. Stock exchanges bring about fund
managers and investment analysts, who monitor and evaluate the industries including the
banking industry. These by-products of stock exchanges provide networking of the
economic sectors and create knowledge and information for the efficient operations of the
markets.
4.7 Conclusions
In an attempt to locate and highlight the role of banking in the economic development, we
discussed in this chapter four models. The first model, section 4.2 forms a basis for the
other models. It takes the bank credit as the banking output that goes into financing the
entrepreneurs’ purchases of capital and labor. It directly links the banking output to the
economic output and productivity. The model does not tell whether or not the banking
output drives the economic output and productivity, but establishes a strong relationship
between the two in the operation of the circular flow of the creditary production. The
second model, section 4.3, is an econometric formulation of Broaddus’ Postulate on
banking efficiency and competition. The banking output that goes into economic
production, as in model one, is a portion of the economy’s savings the banks have been able
to convert to banking output. This process can be inefficient and uncompetitive due to
regulations and banking incompetence. Our econometric formulation takes the regulation as
given and examines the banking competitiveness. Model three, section 4.4, analyzes, in a
comparative static method, the fiscal and monetary policy impacts and the interactive
impacts of the two policies on the output, price level, interest rate, transaction costs and
banking credit. This is done in a Keynesian IS/LM framework with a credit market
assumption that rules out the full information assumption under imperfect credit market
conditions. We also examine the effects of bank credit supply on the other endogenous
variables. The fiscal and monetary policies are found to have no meaningful or definite
impacts on the credit market variables of credit supply, credit demand and the transaction
costs, subsections 4.4.1 and 4.4.2 . Subsection 4.4.3 constructs a relationship between
- 149 -
credit supply and money supply based on the narrow definition of the money. This
relationship shows that money supply is different from the credit supply due to the banking
intermediation and the behavior of the central bank that offers advances to both the
government and the banking industry. This section also models the credit constraint based
on the neo-Keynesian theory of quantity constraint. Credit policy affects the endogenous
variables through the sensitivity of the private investments to the credit quantity and the
banking intermediation factor. If the private investments are insensitive or weakly sensitive
to the bank credit quantity, a situation we denote as non or weak credit constraint
respectively, the credit policy will have smaller impacts on the endogenous variables than
when the credit constraint is strong. With strong credit constraint, increased credit supply
will be followed by increased interest rate; while with weak credit constraint, increased
credit supply will tend to reduce the interest rate. Similarly, under strong credit constraint,
increased credit tends to induce an increase in the demand for credit; while under weak
credit constraint, increased credit supply tends to induce a fall in the demand for the credit.
Fiscal expansion during strong credit constraint will reduce the bank credit supply and
leads to crowding out of the entrepreneurs. A crowding-in effect will be possible if the
credit constraint is very weak. On the contrary, monetary expansion tends to reduce the
bank credit supply during weak credit constraint and tends to increase it during strong
credit constraint. The impact of credit policy and the monetary policy on the output and the
price level are found to coincide if the credit constraint is very weak and the central bank
does not give advances or discounts to the banking system. From section 4.4, we cannot say
for certain that bank credit causes the economic growth or vice versa; this answer will be
attempted in chapter six. Nevertheless, the banking sector cannot be left out in the analysis
of economic growth. In fact, our models have shown that the bank credit stands out among
the other policy variables, its impact on the output is greater than that of the other policy
variables; while its resultant price increase is lower in subsections 4.4.1 and 4.4.2, but
higher in subsection 4.4.3 than that of the other policy variables. This importance of
banking role in the economy is graphically sketched in our modification of Tsuru’s model.
In a model of two economies, bank-based and nonbank-based, the bank-based economy is
found to have higher capital and output levels than the nonbank-based economy. The
presence of banks increases the savings mobilization and specializes savings allocation to
- 150 -
the private sectors. This activity is absent in a nonbank-based economy. However, in the
process of converting the mobilized savings into investments, the banks absorb some
portion of the savings for their own maintenance. High bank absorption means few bank
credit quantities for the economic investments. Thus, the banks become inefficient and
incompetent; the transaction costs of the bank credit will be high resulting in low credit
supply and consequently low investments. In this situation, the banks can retard the
economic growth; the output and capital levels under the bank-based economy could then
be lower than that of the nonbank-based economy. We do not test models for this
comparative analysis in section 4.6. But Schumpeter’s credit in subsection 4.2.1 and section
4.3 model are tested in chapter 5. We also estimate the credit market functions and the
relations between the bank credit and economic activities as modeled in section 4.5. The
four frameworks of theory in this chapter show how complex and ambiguous is the banking
role in the economy. It can be that the bank credit goes directly into the economic
production function, as in model one, section 4.2, or that process is saddled with
regulations and bank incompetence, as in model two, section 4.3, or that with the credit
market imperfections, money-credit relations and the policy shocks, the process can be
promoted or retarded as in model three, section 4.4; or finally that the credit that goes into
the production is actually small fraction of the savings the banks have raised, and the large
fraction is wasted on the maintenance of the banks themselves, as the final model
graphically illustrates, section 4.6. Chapter 5 provides some clues or answers to these
complexities and ambiguities.
- 151 -
5 Analysis of Survey and Empirical Results
5.1 Introduction
In this chapter we analyze the results from the survey and the econometric estimations. The
survey is intended to solicit public views about the banking industry in the Gambia. Its
results will be indicators for the types of conclusions that could be drawn by the
econometric results. The empirical estimations are based on the hypothesized relationships
constructed in the theoretical framework. They will indicate to us the type of hypothetical
relation the banking industry in the Gambia has with the economic activities. We begin
with the survey.
5.2 Survey
In the survey, we attempt to inquire into the following questions:
1. The distribution of account types
2. The reasons for holding a bank account
3. The chance of obtaining a bank facility
4. The reasons for the application being approved
5. The attitude of the public about the banking services
I distributed 150 questionnaires to the public targeting bank customers and the university
students to get informed judgments. The response was 60 % after screening out incomplete
answers. It is hard to retrieve data and information from Gambians. They generally do not
like to accept a responsibility for a source of data or information, and thus they shy away
from any who seeks data and information. Nevertheless, the responses, I have, can mirror to
a certain degree the population’s views. The responses to the five questions hint that
banking industry is not up to the expectations of the public, and that banks give more
weight to consumption financing than to production and entrepreneurial financing. A Chi
squared statistic is computed to test the agreement between the observed distribution and
the theoretical distribution of the responses under each question. The null hypothesis is that
the two distribution are in agreement, and if it is rejected then the two distributions are
significantly divergent. In each of the question above, we take uniform distribution to be
- 152 -
our theoretical distribution, and it means that the responses are uniformly distributed among
the classes in each question; see Yule and Kendall (1950) for details of the test.
5.2.1 Bank Account Distribution
I asked the respondants to list the types of bank accounts they hold. It is important to know
what types of accounts people hold; this explains the bank deposit and credit activities. If
banks can mobilize long time deposits, they will be able to engage in long time credit
commitments to the public. Long time deposit holdings can also indicate the public
confidence in the banks and the national currency. According to our sample, the public
tends to hold current accounts the most followed by savings accounts. Over 50 % of the
respondants hold current accounts compared to just under 40 % of them who hold savings
accounts. None of them holds a time deposit account, and around 28 % of them hold no
bank account. These observed distribution is found to be significantly not in agreement
with the hypothetical uniform distribution, the test-statistic is 20.15. This implies either our
sampling technique is at fault, it does not capture the population correctly, or the theoretical
distribution is at fault. We suspect that the distribution of the observed data closely
represents the population. The distribution of account holdings is not uniformly distributed;
savings and demand deposits account constitute on average more than 60% of the total
funds of the commercial banks, figure 2.6, section 2.4. But Current accounts cannot be used
to mobilize money for project and capital financing. They are only a payment system for
transaction purposes. Thus, high current account holdings signal that the Gambian banks
strive on fees and commissions and short term treasury bills not on dividends and interest
income from project and capital investments. High current account activities are unlikely to
translate into high credit activities for entrepreneurs and project managers.
- 153 -
Figure 5.1 : Account holdings
0
0,1
0,2
0,3
0,4
0,5
0,6
savings time deposit current account others
5.2.2 Reasons for Placing Money in the Banks
We ask our respondants to rank on the order of importance from 1 to 5, with 1 the most
important and 5 the least important, the following reasons for placing money in the banks:
1. To avoid the risk of theft, loss and destruction
2. To earn interest income
3. To qualify for bank facility
4. To make easy payment
5. To save for the future
The higher the number attached to a reason the lesser the importance of that particular
reason. The results show that the Gambian public are concerned first with the easy payment
system they get when they place their money in the banks. The second most important
reason is the safety of money in the banks; that is, they believe that keeping money in the
banks reduce the risk of theft, loss and destruction. The third important reason is to save for
the future. But it is precautionary saving with complete freedom of withdrawal at all times
with minimum loss. Thus, none of them hold time deposits. The other reason for not
holding time deposits is because interest income is the least important reason for the
- 154 -
Gambian people to place their money in the banks. Banks cannot attract the public with
interest rates to mobilize savings, and likewise they cannot attract them with interest rates
to engage in economic activities. In this case, the banks should embark on the universal
banking system, where banks directly undertake economic activities to create entrepreneurs
and companies that will depend on the banks for credit, and the banks will depend on them
for dividends and interest incomes. These conclusions are drawn from the observed
distribution in figure 5.2. The test-statistic is 49.65 and at 1% significance level, it is found
that the observed distribution is divergent from its uniform distribution. Thus, the reasons
for placing money in the banks are significantly different.
Figure 5.2: Reasons for Placing Money in the Banks
0
20
40
60
80
100
120
140
160
risk interest banking facility payment saving
Ran
kin
g
- 155 -
5.2.3 Chance of Obtaining a Bank Facility
We list six financing needs and ask our respondants to rate their chance of obtaining a bank
facility for those needs. The financing needs are,
1. Business/project
2. House purchase/construction
3. Land purchase/development
4. Consumption
5. Agriculture
6. Education
When we exclude consumption from the list above, the other financing needs can be
categorized under investments, and hence comparison can be made between consumption
financing and investment financing. The results show that the consumption has a higher
chance of getting bank facility than the investment. There is 39% chance the respondent’s
application of bank loan for the consumption purpose will be approved. The second highest
chance is for the business/project, while an application of bank loan for land has the lowest
chance of being approved. Thus, the banks tend to favor consumption financing, which
requires only standard operating procedures, than investment financing, which requires
rigorous searching and monitoring. We however find that at 5% significance level the
observed distribution in figure 5.3 is in agreement with its theoretical distribution. The test
statistic is 3.517. This implies that the different purposes / classes in figure 5.3 have equal
chances for the bank loan; for example, consumption financing and capital financing have
equal chances for getting a bank loan.
- 156 -
Figure 5.3: Chance of obtaining bank facility for
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0,45
business/project house p/c land consumption agriculture education
5.2.4 The Reasons for the Credit Application Being Approved
We asked our respondants to tick one of the following securities they thought was the most
relevant in the approval of their credit application:
1. Good Business/project
2. Good collateral
3. Good income flow
4. Good connection in the bank
The respondants indicate that income flow is the most important guarantee for obtaining a
bank credit in the Gambia. Around 34 % of the respondants thought they obtained bank
credits because of their good income flows, which periodically stream into their current
accounts with the banks. 32 % of them thought it was because of the business/project they
had, and 28 % thought it was because of the collaterals they had. Connections and personal
relationships in the banks also play as a factor in obtaining a bank credit. The chi squared
statistic is 7.37 and at 10% significance level, we find that the observed distribution is
divergent from its hypothetical uniform distribution; thus, the differing conclusions from
- 157 -
the observed distribution are significant. It is obvious from the results that people with
good income flows stand a better chance than the rest for getting bank help. These people
are not entrepreneurs, who often do not have steady income flows but ideas and business
proposals that need to be strategically analyzed by the bank management to determine their
viability and profitability. In contrast, to determine one’s eligibility for a bank loan,
particularly for consumption purpose, while one has a steady income streaming into one’s
current account, is a matter of basic financial management for the bank and consumption
smoothing for the applicant. Thus, if this consumption smoothing, which the Gambia banks
often do for their salary accounts and current account holders, can significantly in part
insulate the consumption patterns against adverse shocks, then that practice of banks should
show up in gross domestic product of the country, and the banks shall be counted among
the shock absorbers of the economic growth. This will be ascertained in the empirical
estimation.
Figure 5.4: I have obtained bank facility because I have
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
good project good collateral good income good connection
- 158 -
5.2.5 Public Attitude about the Banking Services
On the Likert scale from 1 to 7, we asked our respondants to rate the banking services in
the Gambia. It appears that the Gambia public are not impressed by the bank performances.
They see the banks as not doing very well as expected. 40 % of the respondants rate the
banking services as just good. 32 % rate the services as satisfactory, and less than 10 % rate
the service as very good. In chapter 2, we explained how the banks operate in the Gambia,
the high clustering of the banks in one area, their tendency of financing few repeated
clients, who are mostly engaged in oligopolistic markets of distributive trades, and hardly
venturing beyond , can explain to some degree why the public have not given the banks a
very good rating, figure 5.5. From the distribution in figure 5.5 we compute a chi squared
statistic of 38.41, and at 1% significance level we find that the observed distribution in
Figure 5.5 is significantly divergent from its uniform distribution. Thus, the sample classes
have significant differences.
Figure 5.5: Ratings of the Banks
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0,45
very poor poor fair satisfactory good very good excellent
- 159 -
5.3 Econometric Analysis In this section, I present and analyze the econometric findings. I first analyze the relations
between the commercial banks and the economic activities; and this covers the goods and
the production markets of our Keynesian model; the empirical results are found in sections
5.3.2, 5.3.3, 5.3.4 and 5.3.5. The economic activities are measured by gross domestic
product, private investments, gross domestic investments, physical capital and economic
efficiency. The banking activities are measured by the liquid liabilities of the banking
industry, the credit volume of the banks to the private sector and the banks’ total assets. I
present the empirical evidence of Broaddus’s model, section 5.3.6, and the analysis of the
commercial bank credits, section 5.3.7, by constructing the supply and demand functions of
the bank credit. The money demand function and the transaction cost function are estimated
to test the hypotheses constructed in chapter 4, section 4.4.3. The analysis of the bank credit
market that includes the money demand function and the transaction costs function covers
the money and credit markets of the Keynesian Model, Chapter 4, section 4.4. I then
investigate the performance of the central bank in terms of the monetary policy competence
as explained in subsection 4.4.4. I have no econometric analysis for Tsuru’s Model,
Chapter 4, section 4.6, due to inaccessible data to make comparison between bank-based
and nonblank-based economy. I begin the analysis with the unit root tests to identify the
integration order of each variable, the test models include intercept and trend. The lag order
is automatic based on Schwarz information criteria, and the maximum lag order is nine.
The VAR order selection test and co-integration rank test results are provided in the
Appendix A. We use Eview 4 software for all the econometric computations.
5.3.1 Unit Root Tests I use the Dickey-Fuller methods to obtain the test results. All the variables in this research
are found to be first difference stationary except the log real liquid liabilities of the
commercial banks, LRLL and the consumer price index, LCPI. These two variables are
found to be second difference stationary. The level variables are not stationary. This
implies that the level regressions may have spurious relations, they reflect coincidental
trending effects of the variables without any economic meaning. Thus, in this research I do
- 160 -
not use the level regressions. I base the analysis on the long run relations estimated using
the Johansen’s co-integration method. This method also gives me an opportunity to analyze
the feed back effects and the weak exogeneity of the variables in the system. The unit root
test reveals that banking variables, liquid liabilities and the bank assets have no significant
trend, while the log credit volume of the banks to the private sector, LDCP, has a trend. In
contrast, all the economic variables exhibit trends. The intercept term is also found
significant in some variables and insignificant in the others. This result motives us to
assume linear deterministic trend in the data and include intercept in the vector auto
regressions. The co-integration tests include intercept and no trend .The table 1 presents the
unit root tests:
Table 1: Unit Root Test Variable Level Fst Difference LGDP
T-statistic -1.904638 -5.983184***
P-value 0.6325 0.0001 LRM
T-statistic -1.599 -7.155***
P-value 0.7746
0.0000
LRPI
T-statistic -2.680 -6.759***
P-value 0.2499 0.0000
LDCP
T-statistic -2.036580 -6.478888***
P-value 0.5632 0.0000
LLC
T-statistic -1.501444 -5.373815***
P-value 0.8116 0.0005
LGDS
T-statistic -3.232053* -4.405894
***
P-value 0.0944 0.0069
LRPK
T-statistic -1.800 -7.019***
P-value 0.6842 0.0000
REFL
T-statistic -2.041 -5.391***
P-value 0.5603 0.000
LCPI
- 161 -
T-statistic -2.482 -2.529
P-value 0.334 0.313
CDC
T-statistic -1.4178 -4.774***
P-value 0.8395 0.002
CNFA
T-statistic -2.015 -3.933**
P-value 0.574 0.020
LRGDP
T-statistic -1.964 -6.872***
P-value 0.602 0.000
LRY
T-statistic -1.937 -6.776***
P-value 0.6157 0.000
LRPLL
T-statistic -2.040 -2.809***
P-value 0.5609 0.000
LRLL
T-statistic -2.024 -2.853
P-value 0.5698 0.1890
LRPDCP
T-statistic -1.765 -6.030***
P-value 0.7013 0.0001
LRDCP
T-statistic -1.756 -6.006***
P-value 0.7057 0.0001
LRPCBA
T-statistic -1.411 -4.692***
P-value 0.8415 0.0030
LRPPI
T-statistic -2.649 -6.819***
P-value 0.2620 0.0000
LRPGDI
T-statistic -1.713 -5.683***
P-value 0.7258 0.0002
RPEF
T-statistic -2.135 -7.145***
P-value 0.5103 0.0000
LRLC
T-statistic -1.385 -4.709***
P-value 0.8492 0.0029
LR/Li
T-statistic -2.682 -7.373***
P-value 0.2490 0.0000
- 162 -
DR
T-statistic -3.236* -7.940
***
P-value 0.0928 0.0000
LRCBA
T-statistic -1.386 -4.710***
P-value 0.8488 0.0029
5.3.2 Gross Domestic Product and the Banking
One co-integration relation exists for the linear combination of GDP and the banking
indicators. We use co-integration lag order of one because its VAR order is two, Appendix
A 1. We also assume a deterministic linear trend in the data, which is an obvious
assumption for the economic variables. This gives us the following unrestricted Vector
Error Correction model in table 2:
Table 2: Output versus Banking (unrestricted model)
*** = significant at 1 %, ** = significant at 5 % * = significant at 10 %
Significant: the variable has no unit root or random walk and thus it is stationary
1 Cointegrating Equation(s): Log likelihood 97.87161
Normalized cointegrating coefficients (std.err. in parentheses) LRY LRPDCP LRPLL LRPCBA 1.000000 4.824430 -2.493148 -0.722662 (0.78209) (0.55990) (0.76913) Adjustment coefficients (std.err. in parentheses) D(LRY) -0.025125 (0.01586) D(LRPDCP) -0.036978 (0.03386) D(LRPLL) 0.120245 (0.03796) D(LRPCBA) 0.007695 (0.02649) LRY = log real per capita output. LRPDCP = log real per capita banks’ credit to the private sector. LRPLL = log real per capita liquid liabilities of the banks. LRPCBA = log real per capita commercial banks’ assets.
- 163 -
The unrestricted co-integration relation has a significant convergence towards the
equilibrium. If the log real per capita GDP increases 1 per cent in the current period above
its equilibrium level, it will subsequently fall by 0.025 per cent in the next period,
indicating that the relation does not have a tendency to explode. Also, the increase in GDP
in the current period spills over to the liquid liabilities of the banks; the liquid liabilities
will increase by 0.12 per cent in the following period following a 1 per cent increase in the
GDP. The bank credit and total asset seem not to receive feedbacks from the level GDP. In
the long run relation, the banks’ assets are found insignificant; while bank credit and liquid
liabilities are significant. Restricting all the adjustment coefficients of the banking
indicators to zero is rejected at 1 per cent significance level. This implies that not all the
banking factors are weakly exogenous. The unrestricted co-integration relation reveals that
bank liabilities are not weakly exogenous; thus we restrict the adjustment coefficients of
bank credit and bank assets to zero, and we get a final reduced model as in table 3.
- 164 -
Table 3: Output versus Banking (restricted model)
We describe this long run relation as output-banking nexus. It is the national output as a
function of the banking development:
( )133.031.129.227.9 ttttt zlrpcdblrplllrpdcplry +++−=
Liquid liabilities, LRPLL, promote the output expansion, LRY, while an increase in the
bank credit supply, LRPDCP contracts the output. This contradicts the hypothesis that bank
credit promotes the output. It means over the period studied, the bank credits in the Gambia
did not go into areas that can promote and expand the national output. The credit activities
were in fact counter productive as the final model indicates. This evidence confirms what
we have earlier noted in the graphical analysis of the banks’ credits versus the major
economic sectors, chapter 2, figure 2.10; there we found that the banks’ credits to the major
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
LRY(-1) 1.000000
LRPDCP(-1) 2.290955 (0.39884) [ 5.74409]
LRPLL(-1) -1.305122 (0.28553) [-4.57084]
LRPCBA(-1) -0.334451 (0.39223) [-0.85269]
C -9.270157
Error Correction: D(LRY) D(LRPDCP) D(LRPLL) D(LRPCBA)
CointEq1 -0.063405 0.229706 (0.03153) (0.04131) [-2.01071] [ 5.56004]
Diagnosis Results: LR Test for the Restrictions χ2(2) = 2.73, Prob. = 0.2556 Autocorrelation Test χ2(32) = 42.78, Prob. = 0.0964 Serial Correlation Test χ2(16) = 21.357, Prob. = 0.165 Normality JB Test χ2(8) = 4.01, Prob. = 0.856 Heterokedasticity Test χ2(200) = 208.158, Prob. = 0.3315
- 165 -
economic activities have been falling. From equation one above, ceteris paribus, 1 per cent
increase in the banks’ credit leads to 2.29 per cent fall in the output. The opposite occurs in
section 5.3.4, equation (7), where bank credit is found to boost the output, we will come to
this later in section 5.3.5. Equation (1) shows that the output will increase by 1.31 per cent
following 1 per cent increase in the liquid liabilities. On the other hand, an increase in the
output causes the economic agents to increase their demand for liquid liabilities to facilitate
the economic transactions that result from the increased output, and thus the liquid
liabilities are endogenous in the model. There is no feedback to the banks’ credit. Banks
create credit to enable the production and in an anticipation of the output sales. This is in
line with the theory of creditary production, chapter 4, section 4.2, where credit is for the
entrepreneur to start the production, and liquid liabilities are the credits of the third parties
to the bank. They demand the liquid liabilities to purchase the output of the entrepreneur.
Thus, banks’ credit to private sector is found to be weakly exogenous with respect to this
equilibrium system, while liquid liabilities are not. Also, pair wise Granger causality has
found that banks’ credit is Granger causal for the gross domestic product , while gross
domestic product is Granger causal for the liquid liabilities, Appendix A 2. Banks’ assets
are found to be Granger non-causal for any variable in the system, and hence its impulses
generate no significant responses from the other variables in the system. A shock to the
banks’ credit creates a significant response from the gross domestic product, and the
response persists on for eight periods, this is found in figure 5.6. Figure 5.6 gives the
impulse responses in the VAR system, where the responses are expected to die out as the
time span increases. Figure 5.7 gives the co-integrated impulse responses, where the
response of a variable to a unit shock is given in the light of the dynamic interaction among
the variables in the system; if the variable has a significant long run coefficient in the
concerned long run equation, its response will be permanent. The responses correspond
closely to the long run coefficients of the variables in the system after taking into account
the dynamic interaction among the variables after a unit shock to one variable. Figure 5.7
gives us further insights about responses of the output to shocks to the banking variables.
Equation (1) shows that a unit shock to the bank credit will result in 2.3 per cent fall in the
output, this is correct assuming that the other variables including the bank credit itself do
not react to the shock. If the other variables, including the bank credit itself, are allowed to
- 166 -
react, then the net response of the output to the unit shock to the bank credit is
approximately 0.03 per cent fall. Similarly, the output will respond by 0.07 per cent
increase to a unit shock to the liquid liabilities; compared to equation (1) that shows that
ceteris paribus the output will increase by 1.3 per cent following a unit shock to the liquid
liabilities. The bank asset has zero impact on the output
- 167 -
Figure 5.6: Impulse-responses of LRY in the VAR System
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRY
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPDCP
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPLL
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPCBA
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRY
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPDCP
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPLL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPCBA
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRY
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPDCP
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPLL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPCBA
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRY
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPDCP
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPLL
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPCBA
Response to Cholesky One S.D. Innovations ± 2 S.E.
- 168 -
Figure 5.7: Impulse-Responses of LRY in the VECM
-.04
-.02
.00
.02
.04
.06
.08
.10
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPDCP
-.04
-.02
.00
.02
.04
.06
.08
.10
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPLL
-.04
-.02
.00
.02
.04
.06
.08
.10
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPCBA
.00
.04
.08
.12
.16
.20
.24
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRY
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRY
.00
.05
.10
.15
.20
.25
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRY
Response to Cholesky One S.D. Innovations
5.3.3 Physical Capital Stock and the Banking
There exists one co-integration relation between the physical capital and the linear
combination of the banking factors, Appendix A3. Liquid liabilities appear irrelevant in this
relation. It is banks’ credit and assets that have significant long run coefficients. The
convergence towards the equilibrium is also significant, indicating that the model is not
explosive. The adjustment coefficients of the banking variables all appear insignificant,
indicating that the banking factors could be weakly exogenous to the equilibrium system.
Table 4 presents the unrestricted co-integration relation.
- 169 -
Table 4: Physical Capital versus Banking (unrestricted model)
It shows that banks’ credit and assets positively improve the capital accumulation in the
economy. Physical capital does not feed back to both banks’ credit and assets, and hence
they are weakly exogenous to the estimation of physical capital. This is revealed in the
following restrictions tests:
1 Cointegrating Equation(s): Log likelihood 80.96239
Normalized cointegrating coefficients (std.err. in parentheses) LRPK LRPDCP LRPLL LRPCBA 1.000000 -0.104135 -0.001226 -0.206137 (0.09157) (0.06713) (0.09639) Adjustment coefficients (std.err. in parentheses) D(LRPK) -0.624580 (0.11539) D(LRPDCP) 0.125029 (0.23543) D(LRPLL) 0.371900 (0.30726) D(LRPCBA) 0.099545 (0.19603)
LRPK = log real per capita physical capital. The other variables are as defined in section 5.3.2
- 170 -
Table 5: Physical capital versus Banking (restricted model)
Ceteris paribus, the contribution of banking development to the capital accumulation is
described by this long run relation,
)2(22.007.019.014.1 ttttt vlrpcbalrplllrpdcplrpk ++−+=
This model supports the claim that banking development increases capital stock
accumulation. Banks like to finance assets that cannot be consumed entirely in one
production cycle in order to provide banks with a sort of collateral. The collaterals are
capital stock for the borrowers; thus, those with high collaterals could expect to have more
bank financing than those with low collaterals. However, the weak exogeneity of the
banking variables implies that it is banking factors that explain changes in the capital stock.
The bank credit and asset go into increasing capital stock. Thus, increased banking
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
LRPK(-1) 1.000000
LRPDCP(-1) -0.194496 (0.09882) [-1.96811]
LRPLL(-1) 0.065601 (0.07245) [ 0.90545]
LRPCBA(-1) -0.217745 (0.10402) [-2.09333]
C -1.135941
Error Correction: D(LRPK) D(LRPDCP) D(LRPLL) D(LRPCBA)
CointEq1 -0.601610 (0.09103) [-6.60913]
Diagnosis Results: LR Test for the Restrictions χ2(3) = 1.343, Prob. = 0.7189 Autocorrelation Test χ2(16) = 16.887, Prob. = 0.393 Serial Correlation Test χ2(16) = 11.67, Prob. = 0.767 Normality JB Test χ2(8) = 5.00, Prob. = 0.757 Heterokedasticity Test χ2(20) = 30.86, Prob. = 0.057
- 171 -
development could cause the diminishing returns to quickly set in due to the resultant
increased capital accumulation. This will take a long time to occur. Our model shows that
banking contribution to the capital accumulation is very small; for the capital stock to
increase by approximately 20 per cent the bank credit or the bank asset has to increase by
100 per cent. This indicates that in the case of the Gambia, the increased banking
development would not cause the diminishing returns to quickly set in. The impulse
responses in figure 5.8 show how long the impacts of banking variables on the physical
capital persist. The physical capital has approximately zero responses to the shocks to the
banking variables; Whereas, the banking variables respond to the shock to the physical
capital. The co-integrated impulse responses in figure 5.9, which correspond to the long run
relation, the physical capital responds significantly to the shocks to the banking variables,
and the banking variables do not respond significantly to the shocks to the physical capital.
A shock to the bank credit will generate 0.035 per cent response from the capital stock, and
similar response will be generated given a shock to the liquid liabilities; this dynamic
interaction shows that the shock to the liquid liabilities causes a positive response from the
physical capital contrary to the result in equation (2), where 1 per cent shock to the liquid
liabilities lead to 0.07 per cent fall in the capital stock. The dynamic interaction in figure
5.9 gives the net effects of the shock, while the long run relation in equation (2) gives the
gross effect of the shock assuming other variables remaining constant.
- 172 -
Figure 5.8: Impulse-Responses of LRPK in VAR
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPK
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPDCP
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPLL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPCBA
-.08
-.04
.00
.04
.08
.12
.16
.20
.24
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPK
-.08
-.04
.00
.04
.08
.12
.16
.20
.24
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPDCP
-.08
-.04
.00
.04
.08
.12
.16
.20
.24
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPLL
-.08
-.04
.00
.04
.08
.12
.16
.20
.24
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPCBA
-.12
-.08
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPK
-.12
-.08
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPDCP
-.12
-.08
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPLL
-.12
-.08
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPCBA
-.04
.00
.04
.08
.12
.16
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPK
-.04
.00
.04
.08
.12
.16
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPDCP
-.04
.00
.04
.08
.12
.16
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPLL
-.04
.00
.04
.08
.12
.16
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPCBA
Response to Cholesky One S.D. Innovations ± 2 S.E.
- 173 -
Figure 5.9: Impulse-Responses of LRPK in a VECM
.00
.01
.02
.03
.04
.05
.06
.07
.08
.09
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPK
.00
.01
.02
.03
.04
.05
.06
.07
.08
.09
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPDCP
.00
.01
.02
.03
.04
.05
.06
.07
.08
.09
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPLL
.00
.01
.02
.03
.04
.05
.06
.07
.08
.09
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPCBA
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPK
-.05
.00
.05
.10
.15
.20
.25
.30
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPK
.00
.02
.04
.06
.08
.10
.12
.14
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPK
Response to Cholesky One S.D. Innovations
- 174 -
5.3.4 Economic Efficiency and the Banking
We find one co-integration relation between the economic efficiency variable and the linear
combination of the banking factors, Appendix A6. It is unexpected to find banks’ credit
irrelevant in this system. Credit activities of banks are expected to co-move with the
economic efficiency. If bank credit activity is associated with reduced transactions costs
and improved information flow, it should have a significant long run coefficient in this
system. But as both the unrestricted relation in table 6 and the restricted relation in table 7
show the log real per capita banks’ credit to the private sector is irrelevant in this system :
Table 6: Economic Efficiency versus Banking (unrestricted model)
The evidence here indicates that the credit activities of the banks in the Gambia do not
contribute significantly to reduced transactions costs and improvements in the information
flow. This is augmented by another evidence that no long run relation exists between the
investments, either private or gross domestic investment, and the linear combination of the
banking factors alone. The total banks’ credit to the major economic activities have been
declining, and the average lending rate has been soaring up with the spread between
lending rates and the deposit rates ever widening, and this spread proxies for the transaction
1 Cointegrating Equation(s): Log likelihood 145.3177
Normalized cointegrating coefficients (std.err. in parentheses) REFL LRDCP LRLL LRCBA 1.000000 0.034525 -0.052716 0.110678 (0.04649) (0.02090) (0.03237) Adjustment coefficients (std.err. in parentheses) D(REFL) -0.339193 (0.03059) D(LRDCP) 0.142378 (0.32398) D(LRLL) 0.322561 (0.43058) D(LRCBA) 0.136503 (0.27081) REFL = log real economic efficiency per labor force LRDCP = log banks’ credit to the private sector. LRLL = log real liquid liabilities of the banks. LRCBA = log real commercial banks’ assets.
- 175 -
costs; and this implies the lending activity has been the transaction costs. We will come to
this point in Broaddus’ model, section 5.3.6.
Table 7: Economic Efficiency versus Banking (restricted model)
The economic efficiency as a function of the banking development,
)3(11.005.003.005.3 ttttt ulrcbalrlllrdcprefl +−+−=
The evidence in section is in contrast with the increasing number of banks in the country;
increasing number of banks should have induced competition, and as a result the gap
between the lending rates and the deposit rates would have been dwindling, and the
information flow should have been increasing and promoting efficiency. The fact that
banks’ credit to the major economic sectors have been shrinking indicates that the nation’s
scarce savings resources have been inefficiently allocated. Therefore, the credit activities of
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
REFL(-1) 1.000000
LRDCP(-1) 0.030123 (0.04587) [ 0.65676]
LRLL(-1) -0.048164 (0.02062) [-2.33546]
LRCBA(-1) 0.111297 (0.03194) [ 3.48435]
C -3.048514
Error Correction: D(REFL) D(LRDCP) D(LRLL) D(LRCBA)
CointEq1 -0.337932 0.000000 0.000000 0.000000 (0.02506) (0.00000) (0.00000) (0.00000)
Diagnosis results: LR Test for Restriction χ2(3) = 0.619, Prob. = 0.892 Autocorrelation Test χ2(80) = 76.389, Prob. = 0.594 Serial Correlation Test χ2(16) = 11.67, Prob. = 0.766 Normality JB Test χ2(8) = 36.174, Prob. = 0.000 Heterokedasticity Test χ2(20) = 37.70, Prob. = 0.01
- 176 -
the banks are irrelevant for the efficiency estimation. The banks’ liquid liabilities and assets
are significant in the long run. The use of liabilities as means of payment promote
efficiency in the economy. The banks’ assets are inefficiently allocated, and as a result
increased holding of misallocated assets can only reduce the economic efficiency.
The restrictions, in table 7, that the banking variables are weakly exogenous with respect to
the equilibrium system, which is normalized by the long run coefficient of the economic
efficiency, cannot be rejected. This presents another evidence that banking development is
not endogenous to the economic development. Changes in the banking activities
significantly lead changes in the economic factors. The impulse responses in figure 5.10
show that the economic efficiency does not react to the shocks to the bank liquid liabilities
but negatively react to the shock to bank assets. It responds positively to the shock to the
bank credit. The responses do die out quickly. The banking variables respond
insignificantly to the shocks to the economic efficiency. These results confirm that of
equation (3), where the banking variables are found not receiving feedbacks from the
economic efficiency, and that the bank asset has a negative association with the economic
efficiency. The effects of shocks on the economic efficiency to the banking variables in
figure 5.11 are negative. That is, in a dynamic interaction among the variables in the
system, the economic efficiency will fall following a shock to any banking variable in the
system. On the other hand, a shock to the economic efficiency has zero impacts on the
banking variables, further strengthening the weak exogeneity of these variables in the
system.
In the next section, we expand our model to consists of four economic factors and the three
banking variables. This gives us an opportunity to analyze the importance of the banking
activities in the presence of other economic variables that can operate in the system as
control variables.
.
- 177 -
Figure 5.10: Impulse-Responses of REFL in the VAR system
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRLL to REFL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRLL to LRDCP
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRLL to LRLL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRLL to LRCBA
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRDCP to REFL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRDCP to LRDCP
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRDCP to LRLL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRDCP to LRCBA
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRCBA to REFL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRCBA to LRDCP
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRCBA to LRLL
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LRCBA to LRCBA
-.03
-.02
-.01
.00
.01
.02
.03
1 2 3 4 5 6 7 8 9 10
Response of REFL to REFL
-.03
-.02
-.01
.00
.01
.02
.03
1 2 3 4 5 6 7 8 9 10
Response of REFL to LRDCP
-.03
-.02
-.01
.00
.01
.02
.03
1 2 3 4 5 6 7 8 9 10
Response of REFL to LRLL
-.03
-.02
-.01
.00
.01
.02
.03
1 2 3 4 5 6 7 8 9 10
Response of REFL to LRCBA
Response to Cholesky One S.D. Innovations ± 2 S.E.
- 178 -
Figure 5.11: Impulse-Responses of REFL in VECM
-.015
-.010
-.005
.000
.005
.010
.015
.020
1 2 3 4 5 6 7 8 9 10
Response of REFL to REFL
-.015
-.010
-.005
.000
.005
.010
.015
.020
1 2 3 4 5 6 7 8 9 10
Response of REFL to LRDCP
-.015
-.010
-.005
.000
.005
.010
.015
.020
1 2 3 4 5 6 7 8 9 10
Response of REFL to LRLL
-.015
-.010
-.005
.000
.005
.010
.015
.020
1 2 3 4 5 6 7 8 9 10
Response of REFL to LRCBA
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRDCP to REFL
.00
.05
.10
.15
.20
.25
.30
1 2 3 4 5 6 7 8 9 10
Response of LRLL to REFL
.00
.02
.04
.06
.08
.10
.12
.14
1 2 3 4 5 6 7 8 9 10
Response of LRCBA to REFL
Response to Cholesky One S.D. Innovations
- 179 -
5.3.5 Economic Activities and the Banking
We place all the economic variables, except the gross domestic investments, with the
banking variables in one vector error correction model. We exclude the gross domestic
investments to reduce the dimension of the system and economize the degrees of freedom
of the system. Also, the gross domestic investments and private investments are each
separately not found to be co-integrating with the linear combination of the banking
variables, Appendices A4 and A5. The banks’ credit to the private sector is expected to co-
integrate with the private investments than with the gross domestic investments that include
government investments; and hence we maintain the private investments and exclude the
gross domestic investments from the system in this section. The economic variables are log
real per capita GDP, log real per capita private investments, log real per capita physical
capital and the economic efficiency. These variables are then followed in an ad hoc
ordering by log real per capita banks’ credit to the private sector, log real per capita liquid
liabilities of the banks and log real per capita banks’ assets. We find three co-integration
relations in this system, appendix A7. Table 8 presents the unrestricted systems, the
automatic restrictions there on the long run coefficients are needed to the make the system
just identified we perform the economic restrictions in table 9. The automatic restrictions
however shows that each co-integrating row is normalized by an economic variable. Row
one is the co-integration relation between the log real per capita GDP and the linear
combination of the other variables in the system, row two is the co-integration relation
between the log real per capita private investments and the linear combination of other
variables in the system excluding the log real per capita GDP and the log real per capita
capital; and row three is the co-integration relation between the log real per capita physical
capital and other variables in the system excluding log real per capita GDP and log real per
capita private investments. Banks’ credit is significant only in the second and the third co-
integrating rows; it is irrelevant in the estimation of log real per capita GDP. Also, the log
real per capita liquid liabilities are irrelevant in the estimation of log real per capita GDP.
Banks’ assets are irrelevant in the second co-integration row but relevant in the first and
the third co-integration rows. Improvements in the real per capita economic efficiency has
almost one-to-one relation with the log real per capita GDP, the log real per capita private
investments and the log real per capita physical capital. Efficiency substantially contributes
- 180 -
to the economic growth. Banks’ credit and liquid liabilities have insignificant contributions
to the real per capita GDP in the presence of economic efficiency and bank assets.
Furthermore, an increase in the banks’ credit constitutes falls in both log real per capita
private investments and log real per capita physical capital. That is, banks’ credits in the
Gambia do not promote entrepreneurs or capital formation. They are directed to personal
loans of consumptions and distributive trades of import and export. Project and
development financings are seldom undertaken by the Gambian banks. We notice this fact
in chapter 2, figure 2.10, where their portfolios in the distributive trades and foreign assets
have been expanding. Liquid liabilities are relevant in the second and the third co-
integration equations, and they positively contribute to the log real per capita private
investments and the log real per capita physical capital. This could be accrued to the
payment facility the liquid liabilities provide for the economic agents. In the survey, we
found that easy payment is one of the most important reasons why the economic agents
place their money in the banks. The unrestricted findings of the system are presented in
table 8.
- 181 -
Table 8: Economic Activities versus Banking (unrestricted model)
If one has demand deposit accounts with banks one will be able to make easy payments
through the banks to purchase capital and investment goods; but one cannot rely on the
bank credits to finance the purchases of capital and investment goods. The other reason for
the negative contribution of banks’ credit to both private investments and physical capital
accumulation could be due to the evidence derived in section 5.3.4, where we find that
banks’ credit does not induce efficiency. The credits are not efficiently allocated, and hence
they cannot promote investments and capital formation. Table 8 has an assumption that
each co-integration equation represents a behavioral relationship between one economic
3 Cointegrating Equation(s): Log likelihood
296.8835
Normalized cointegrating coefficients (std.err. in parentheses) LRY LRPPI LRPK RPEF LRPDCP LRPLL LRPCBA 1.000000 0.000000 0.000000 -1.030812 -0.001150 0.007162 -0.084116 (0.05303) (0.03075) (0.02056) (0.02965) 0.000000 1.000000 0.000000 -0.945896 0.317793 -0.104399 -0.156870 (0.21030) (0.12193) (0.08153) (0.11757) 0.000000 0.000000 1.000000 -0.670520 0.407573 -0.139964 -0.292735 (0.15457) (0.08962) (0.05993) (0.08641) Adjustment coefficients (std.err. in parentheses) D(LRY) -0.180450 -0.002536 -0.021854 (0.24986) (0.12696) (0.12702) D(LRPPI) 0.113953 -0.607388 -0.105502 (0.38757) (0.19692) (0.19702) D(LRPK) -0.769334 0.292427 -0.455704 (0.16874) (0.08574) (0.08578) D(RPEF) 0.280502 -0.069222 -0.026430 (0.23538) (0.11960) (0.11966) D(LRPDCP) -0.074186 0.190855 -0.293877 (0.46295) (0.23522) (0.23533) D(LRPLL) -0.928054 0.372881 0.539988 (0.57838) (0.29388) (0.29401) D(LRPCBA) -0.250150 0.132268 0.026395 (0.39313) (0.19975) (0.19985)
LRY = log real per capita GDP. LRPPI = log real per capita private investments. LRPK = log real per capita physical capital. RPEF = log real per capita economic efficiency. LRPDCP = log real per capita banks’ credit to the private sector. LRPLL = log real per capita liquid liabilities of the commercial banks. LRPCBA = log real per capita commercial banks’ assets.
- 182 -
factor and the linear combination of the banking variables including the economic
efficiency. For example, equation four assumes that in the long run output co-move only
with economic efficiency and the banking variables, the private investment and the capital
are set to zero. The assumption though unrealistic helps us examine the behavior of banking
in the presence and in the absence of relevant economic factors. In table 9, we relax this
assumption and test also the weak exogeneity of the banking variables. We normalize
equation four by the long run coefficient of the output and set long run coefficient of bank
asset to zero. The bank asset was found insignificant in the individual model of the output,
section 5.3.2. We assume that goods market (here output model) does not require bank
assets but bank credit and liquid liabilities to facilitate the transactions of the goods; we
also set the long run coefficient of the economic efficiency to zero in equation four. In
subsection 4.4.4, we derived the economic efficiency from a production function using the
output and the capital variables, and thus the presence of the output and the capital in
equation four motivates us to remove the economic efficiency variable to improve over a
possible problem of autocorrelation. Similarly, the same reason motivates us to remove
economic efficiency variable from equation six because both the output and the capital are
present there. Equation five is normalized by the long run coefficient of private investment.
In this equation we set both the long run coefficients of the output and the capital to zero on
the assumption that the economic efficiency variable sufficiently represents both the output
and the capital. We restrict the long run coefficient of liquid liabilities to zero in equation
six; in the individual model of capital accumulation, the liquid liabilities are found
insignificant and this motivates us to set it zero in equation six. Finally, we set the
adjustment coefficients of the economic efficiency and banking variables to zero to test
their weak exogeneity in the system. Table 9 presents the results.
- 183 -
Table 9: Economic Activities versus Banking (restricted model)
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq:
CointEq1 CointEq2 CointEq3
LRY(-1) 1 - 0.365549 (0.09010) [ 4.05699]
LRPPI(-1) -1.238554 1 -1.001170 (0.06468) (0.10676) [-19.1477] [-9.37748]
LRPK(-1) 0.446325 - 1 (0.02342) [ 19.0568]
RPEF(-1) - -1.114093 - (0.08100) [-13.7545]
LRPDCP(-1) -0.194588 0.143844 -0.054208 (0.08475) (0.10602) (0.07201) [-2.29614] [ 1.35682] [-0.75280]
LRPLL(-1) 0.011465 0.036810 - (0.03550) (0.06515) [ 0.32298] [ 0.56504]
LRPCBA(-1) - -0.137280 -0.108017 (0.05361) (0.04840) [-2.56087] [-2.23181]
C 3.951162 -3.926807 2.851963
Error Correction:
D(LRY) D(LRPPI) D(LRPK) D(RPEF) D(LRPDCP) D(LRPLL) D(LRPCBA)
CointEq1 -0.525545 0.953305 -0.340188 - - - - (0.04085) (0.43347) (0.22436) [-12.8649] [ 2.19923] [-1.51625]
CointEq2 -0.343725 -0.276103 -0.502873 - - - - (0.02072) (0.21991) (0.11382) [-16.5851] [-1.25551] [-4.41797]
CointEq3 0.246183 -0.786612 -0.364047 - - - - (0.03209) (0.34051) (0.17625) [ 7.67150] [-2.31007] [-2.06555]
Diagnosis results: LR Test for the Restrictions χ2(12) = 13.166, Prob. = 0.357 Autocorrelation Test χ2(245) = 270.14, Prob. = 0.130 Serial Correlation Test χ2(49) = 56.69, Prob. = 0.21 Heterokedasticity Test χ2(168) = 178.51, Prob. = 0.275
- 184 -
The restrictions give the following long run system of equations:
( )401.019.045.024.195.3 tttttt elrplllrpdcplrpklrppilry +−+−+−=
( )514.004.014.011.193.3 tttttt vlrpcbalrplllrpdcprpeflrppi ++−−+=
( )611.005.000.137.085.2 tttttt zlrpcbalrpdcplrppilrylrpk ++++−−=
The long run system of equations reveals that the importance of banking erodes in the
presence of relevant economic variables. Equation four is the output function that depends
on the private investment, capital stock, banks’ credit to private sector and the liquid
liabilities of the banks. The long run coefficients of the private investment, the capital stock
and the bank credit are significantly different from zero, while the liquid liabilities variable
has zero impact on the output. The banking variables in this equation have the expected
signs, increased bank credit is associated with increased trades, and as a result the output
must increase following an increase in the bank credit. In the individual model of output
function, section 5.3.2, bank credit has a negative sign indicating that individual models
could produce misleading results due to the absence of control variables. Equation five is
the private investment function that depends on the economic efficiency, bank credit to the
private sector, liquid liabilities and the bank asset. There is approximately one-to-one
relation between the economic efficiency and the private investment. Efficiency resulting
from reduced transaction costs, increased information flow or improved method of
production will definitely increase the private investments. Also, increased bank credit
should have a positive impact on the private investment. But the result, though significant
only at 10 %, shows that the bank credit increase reduces the private investments. This
could be true, if the banks that have the nation’s savings choose to place the funds in only
distributive trade, import and export, and ignore to incubate financially the local industries,
or help develop the local entrepreneurs, then increased bank credit may not have a telling
impact on the private sector activities. However, increased bank asset increases the private
investments; this could result from the fact that banks’ asset trading involves the private
sector investments. The private sector also can infer some relevant information from the
types of assets the banks hold and hence motivating the private sector to invest in those
assets because the banks are believed to have superior information. Equation six is a capital
accumulation function that depends on the output, the private investment, the bank credit
- 185 -
and the bank assets. The result shows one-to-one relation between the private investment
and the capital accumulation. The bank credit is irrelevant to the capital accumulation,
while increased bank asset promotes capital accumulation. The system of equations shows
that liabilities side of the bank balance sheet is irrelevant as the economic activities are
concerned. The two variables from the asset side of the balance sheet, bank credit and bank
asset, are proven relevant. Bank credit positively facilitates the trading of the output; while
the bank asset positively promotes the private investments and the capital accumulation. It
also shows that bank impact on the economic activities are eroded and in some cases
rendered irrelevant in the presence of control variables. The impact of banks is more visible
in the individual models, where the control variables are absent. On the exogeneity of the
banking factors, the test reveals that all the banking variables are weakly exogenous with
respect to the equilibrium system. This means banking factors are explanatory variables for
the economic activities; and that the banking factors feed back onto the economic activities,
while the economic activities do not feed back onto the banking factors, except in the
individual models, where we find that output does feed back onto the liquid liabilities of the
banks.
Four long run coefficients are found insignificant in table 9. Among the banking variables it
is only the banks’ assets that are found relevant in the private investments function. The
private investments are the only economic variable in table 9 that are found not co-
integrating with the banking variables, appendix A4, then they may not have a long run
function as given in equation 5. Also, the presence of the private investments has made the
set of the economic variable to have a full rank, Appendix A11. The set that contains only
the output, the physical capital and the economic efficiency has no co-integration rank,
Appendix A12. Likewise, the set of the banking variables has no co-integration rank,
Appendix A13. Thus, the three co-integrating relations in table 9 could be corresponding
closely to the combination of output and the banking variables, section 5.3.2, capital and
banking variables, section 5.3.3 and economic efficiency and banking variables, section
5.3.4. We therefore remove the private investments and re-estimate section 5.3.5 model.
Now both the two sets, the set of the economic variables and the set of the banking
variables, have no co-integration relation, Appendices A12 and A13. After combining them
into one set we get three co-integrating relations, Appendix A8; and the weak exogeneity of
- 186 -
all the banking variables in the system is rejected, Appendix A9. Table 10 below presents
the unrestricted system, and table 11 presents the system with the binding restrictions.
Table 10: Economic Activities versus the Banking (unrestricted system, six variables)
In the unrestricted system above, many adjustment coefficients of the banking variables
appear insignificant, but restricting all of them to zero is rejected at 1 % significance level,
Appendix A 9. This forces us to evaluate their weak exogeneity separately, and we then
find that only the adjustment coefficients of LRPDCP in columns one and two can be
significantly restricted to zero. The rationale for the restrictions on the long run coefficients
to make the system just identified are the same like that given for the table 9 restrictions on
the long run coefficients. Table 11 below gives the restricted system:
3 Cointegrating Equation(s): Log likelihood 347.2156
Normalized cointegrating coefficients (std.err. in parentheses) LRY LRPK RPEF LRPDCP LRPLL LRPCBA
1.000000 0.000000 0.000000 -0.706894 0.122242 0.214646 (0.15668) (0.08359) (0.08417)
0.000000 1.000000 0.000000 0.592911 -0.371343 -0.486683 (0.12961) (0.06915) (0.06963)
0.000000 0.000000 1.000000 -1.081611 0.297787 0.483034 (0.21368) (0.11400) (0.11479)
Adjustment coefficients (std.err. in parentheses) D(LRY) -0.233874 -0.404836 -0.187943
(0.66036) (0.22236) (0.54179) D(LRPK) 0.777567 -0.661730 -0.655213
(0.98728) (0.33244) (0.81001) D(RPEF) -0.082256 -0.413138 -0.309462
(0.54310) (0.18288) (0.44559) D(LRPDCP) 1.018557 0.303126 -0.083805
(2.30845) (0.77731) (1.89396) D(LRPLL) 6.003241 -0.497402 -4.757040
(2.55748) (0.86116) (2.09828) D(LRPCBA) 2.424696 0.452361 -1.895168
(1.66786) (0.56161) (1.36839) LRY = log real per capita GDP. LRPK = log real per capita physical capital. RPEF = log real per capita economic efficiency. LRPDCP = log real per capita banks’ credit to the private sector. LRPLL = log real per capita liquid liabilities of the commercial banks. LRPCBA = log real per capita commercial banks’ assets
- 187 -
Table 11: Economic Activities versus Banking (restricted system, six variables) All the long run coefficients in table 11 are significantly different from zero, while table 9
has four insignificant long run coefficients. Banking role has dwindled in table 9, while in
table 11 the banks’ credit has maintained the expected sign in all the three equations, it
shows that banks’ credit significantly promote output, capital accumulation and economic
efficiency.
Cointegrating Eq:
CointEq1 CointEq2 CointEq3
LRY(-1) 1.000000 3.037934 0.000000 (0.02228) [ 136.377]
LRPK(-1) 0.441088 1.000000 0.000000 (0.00871) [ 50.6324]
RPEF(-1) 0.000000 0.000000 1.000000
LRPDCP(-1) -0.440969 -1.546167 -1.093007 (0.04428) (0.12880) (0.12015) [-9.95932] [-12.0046] [-9.09670]
LRPLL(-1) -0.042043 0.000000 0.302554 (0.00757) (0.05699) [-5.55537] [ 5.30910]
LRPCBA(-1) 0.000000 0.165698 0.483709 (0.02398) (0.06470) [ 6.90907] [ 7.47561]
C -1.632926 -3.982628 1.304955
Error Correction: D(LRY) D(LRPK) D(RPEF) D(LRPDCP) D(LRPLL) D(LRPCBA)
CointEq1 -2.904254 -8.111217 -3.432263 0.000000 -21.80063 -2.698942 (3.46796) (5.19895) (2.99072) (0.00000) (11.5746) (5.75171) [-0.83745] [-1.56016] [-1.14764] [ NA ] [-1.88349] [-0.46924]
CointEq2 0.843579 2.869998 1.085166 0.000000 8.926404 1.477616 (1.33422) (2.00018) (1.15061) (0.00000) (4.45306) (2.21284) [ 0.63227] [ 1.43487] [ 0.94312] [ NA ] [ 2.00456] [ 0.66775]
CointEq3 -0.109007 -0.530316 -0.269712 0.651059 -4.263347 -1.434025 (0.50305) (0.75410) (0.43333) (0.28402) (1.68682) (0.84907) [-0.21669] [-0.70324] [-0.62242] [ 2.29231] [-2.52744] [-1.68894]
Diagnosis results: LR Test for the Restrictions χ2(2) = 2.973, Prob. = 0.226 Serial Correlation Test χ2(36) = 41.683, Prob. = 0.237 Heterokedasticity Test χ2(630) = 640.6, Prob. = 0.371
- 188 -
The long run system of equations of table 11 are,
( )7042.0441.044.0634.1 ttttt eLRPLLLRPDCPLRPKLRY +++−=
( )8166.055.104.3983.3 ttttt uLRPCBALRPDCPLRYLRPK +−+−=
( )9484.0302.0093.13.1 ttttt vLRPCBALRPLLLRPDCPRPEF +−−+−=
Equation (7) shows that log real per capita banks’ credit to the private sector, LRPDCP, and
the log real per capita banks’ liquid liabilities, LRPLL, have positive impacts on the log
real per capita GDP, LRY, and the log real per capita physical capital has a negative impact
on LRY. Increased physical capital accumulation accelerates the diminishing returns to
scale, and hence in the long run capital accumulation will depress the output growth.
Section 3.3.4 gives empirical evidence that banks retard the economic growth for the fact
that their financing activities tend to promote capital accumulation. Equation (7) does not
support this evidence, the banks’ credit here promotes the output and does not retard its
growth. Equation (8) does provide evidence that banks’ credit promotes capital
accumulation, ceteris paribus, 1 per cent increase in the banks’ credit, LRPDCP, will lead
to 1.55 per cent increase in LRPK. If the banks credit retards economic growth by
promoting capital accumulation, then the long run coefficient of bank credit in equation (7)
should have been negative and positive in equation (8). Banks’ credit is also found to
improve the economic efficiency, while the banks’ assets and liquid liabilities do not,
equation (9). The co-integrated impulse responses in figure 6.13 also reveal that LRY reacts
positively to LRPDCP shock and LRPLL shock as equation (7) states. A shock to LRPCBA
generates almost a zero response from the LRY, LRPK and RPEF. This shows that in a
dynamic interaction, the importance of the banks’ assets becomes almost negligible; while
the banks’ credit and liquid liabilities have persistent impacts. This outcome is close to the
results of the Granger causality test, Appendix A10. In this test, the banks’ credit and liquid
liabilities are found to be Granger causal for all the variables in the system except the
physical capital, and none of the variables is Granger causal for them. The economic
variables are non-Granger causal for the banking variables. Thus banking activity has direct
impact on the economy; its increased credit activity and liquid liabilities have positive long
run impacts on the output growth. The increased bank credit to the private sector also
increases the capital accumulation and improves the economic efficiency. But the increased
- 189 -
total asset activity of the banks tends to reduce both capital accumulation and economic
efficiency; similarly the increased liquid liabilities are not efficiency improving.
- 190 -
Figure 5.12: VEC Impulse responses, Economic variables versus banking
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRY
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPDCP
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPLL
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRY to LRPCBA
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPK
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPDCP
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPLL
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LRPK to LRPCBA
-.10
-.05
.00
.05
.10
1 2 3 4 5 6 7 8 9 10
Response of RPEF to RPEF
-.10
-.05
.00
.05
.10
1 2 3 4 5 6 7 8 9 10
Response of RPEF to LRPDCP
-.10
-.05
.00
.05
.10
1 2 3 4 5 6 7 8 9 10
Response of RPEF to LRPLL
-.10
-.05
.00
.05
.10
1 2 3 4 5 6 7 8 9 10
Response of RPEF to LRPCBA
-.10
-.05
.00
.05
.10
.15
.20
.25
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRY
-.10
-.05
.00
.05
.10
.15
.20
.25
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPK
-.10
-.05
.00
.05
.10
.15
.20
.25
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to RPEF
-.10
-.05
.00
.05
.10
.15
.20
.25
1 2 3 4 5 6 7 8 9 10
Response of LRPDCP to LRPDCP
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRY
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPK
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to RPEF
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LRPLL to LRPLL
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRY
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPK
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to RPEF
-.04
.00
.04
.08
.12
.16
.20
1 2 3 4 5 6 7 8 9 10
Response of LRPCBA to LRPCBA
Response to Cholesky One S.D. Innovations
- 191 -
5.3.6 Broaddus’ Model
Competitive banking industry can efficiently allocate the nation’s savings, while a
monopolist system will breed inefficiency, and the scarce savings will be inefficiently
allocated. The activity of efficient allocation of savings will “hold the loan interest rate at
exactly the level required to equate total revenue and total deposit interest costs”, Broaddus
(1973). As the model is presented in chapter four, the loan interest rate and the deposit
interest rate must move in an equilibrium relationship. At 5% significance level under both
trace and max-eigenvalue tests we have one co-integration relation, Appendix A14. The
loan interest rate and the deposit interest rate are found to co-move in an equilibrium
relationship. Thus, the existence of this long run relation rules out the existence of spurious
relation between the loan interest rate and the deposit interest rate. The loan interest rate is
found to feed back to the deposit interest rate. A 1 per cent increase in the loan interest rate
tends to increase the deposit interest rate by 0.3 per cent in the subsequent period. Table 12
gives us the results.
Table 12: Lending and Deposit rates
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
LR(-1) 1.000000
DR(-1) -3.058189 (0.28625) [-10.6838]
C 11.98716
Error Correction: D(LR) D(DR)
CointEq1 0.462225 0.286881 (0.11148) (0.07066) [ 4.14608] [ 4.06017]
LR = lending interest rate. DR = deposit interest rate. Diagnosis results: Autocorrelation Test χ2(16) = 17.88, Prob. = 0.331 Serial Correlation Test χ2(4) = 3.265, Prob. = 0.514
- 192 -
Broaddus’ Model,
( )10058.3987.11 ttt eDRLR ++−=
where the estimated parameters are,
67.0,
1
1058.3
987.11
=−
==
−=
D
D
rrthen
rrβ
α
The intercept term is significantly different from zero. This indicates that the banking
industry is not operating at an efficient level. The inefficiency represents case one in
chapter 4. It is the case where the macroeconomic policies suppress lending rates to achieve
a positive shift of aggregate demand, the lending rates are repressed. The negative intercept
implies that the lending rate is below its equilibrium level, and the borrowers have been
subsidized. Also, the empirical reserve ratio of 67% is substantially larger than our
calculated average regulatory reserve ratio of 19.27% over the observation period. This
means the banking industry has deviated substantially from the equilibrium point, where
the empirical reserve ratio is equal to the regulatory reserve ratio; and it is manifest in the
large spread between the lending interest rate and the deposit interest rate, and as a results
the banking industry is uncompetitive and inefficient. This empirical evidence is very close
to the banking realities in the Gambia in the 1960’s, 1970’s and early 1980’s. In these
periods, the dominant banks in the Gambia were Gambia Agricultural Development Bank
and Gambia Commerce and Development Bank. The two banks lent huge amounts of loans
at below market interest rates resulting in an inefficient allocation of the savings and misuse
of the loan facilities. They eventually went bankrupt, and the assets of one of them is still
being recovered by Gambia Assets and Management Recovery Corporation, a company
that was set up purposely for that. At the time it went bankrupt, the Agricultural
Development Bank had 44% of the total liabilities of banking industry. Thus, the evidence
here that the lending interest rate has been repressed, the loan takers have been subsidized
and that the banking industry is inefficient and uncompetitive given the large empirical
reserve ratio cannot be rejected. In the next section, I analyze the loans market of the
- 193 -
banking industry and attempt to identify the credit demand and credit supply functions.
This will help us identify the important determinants of the two functions.
5.3.7 Banks’ Credit Market
I explore here the demand for and the supply of banks’ credit from supply and demand
views. In addition to identifying the determinants of the two functions, I can characterize
the shapes of the two curves based on the estimated slope coefficients and determine
whether or not the bank credit is supply constrained. The functions are taken from Catao
(1997) as follows:
( )ttt LCLRfS ,= : Supply function of banks’ credit
=
ttttt GDP
GDPDGDPLRfD ,,, : Demand function of banks’ credit
tt DS = : at the equilibrium
where,
S = credit volume of the banks
D = credit demand
LR = lending rate
LC = lending capacity of the banking: demand deposits + capital
PD = private sector debt.
GD = government current expenditure
GDP = gross domestic product.
The demand for the banks’ credit is determined by the costs of borrowing, (LR) and by the
economic activity. I drop out the measures of private sector debt and the government
expenditure on the assumption that the lending rate has already embodied these variables,.
The supply and the demand together determine the equilibrium. Thus my vector auto-
regression, z, consists of the variables as below:
( )ttttt GDPLCDRLRDCPz ,,,,=
The deposit interest rate, DR, measures here the maintenance costs of the deposits, and we
use it also to measure the profitability on the demand side. The nominal GDP measures the
- 194 -
products in terms of the nominal prices. Changes in GDP could be due to nominal price
changes or real changes in the economic activities or both. Thus, Banks’ credit supply
depends on the lending interest rate, LR, the cost of deposits, DR, the lending capacity, LC,
GDP and the level of economic activities, GDP. Problem loans are important factors in the
credit supply of the banks, it is only that no reliable data could be collected on this variable
in the Gambia.
Supply and Demand Functions of Bank Credit
Under the assumption of linear deterministic trend in the data, two co-integration relations
are estimated at both 1 % and 5 % significance levels by both trace test and the max-
eigenvalue test, Appendix A15. To identify these relations as supply and demand functions,
I introduce certain restrictions. The first co-integration row in table 13 is taken as the
supply curve of bank credits, and it is normalized by the long run coefficient of LDCP. The
second co-integration row is taken as the inverse demand function of the bank credits, and
it is normalized by the long run coefficient of LR. In the supply function, I restrict the
coefficient of the LGDP to be zero, and in the demand function, I restrict the coefficient of
the LLC to be zero. In other words, the effect of GDP in the credit supply function is
captured by the lending capacity. Similarly, the effect of the lending capacity is captured
by the GDP and the profitability measure of DR in the entrepreneurs’ demand for bank
credit. Deposit interest rates are costs to the bankers but returns on the entrepreneurs’
savings. They are relevant in the determination of both demand and supply functions of
credit. With these identification restrictions, only the deposit interest rate is found to be
weakly exogenous with respect to the co-integration equation one, the supply function.
Table 14 gives the restricted results.
- 195 -
Table13: Demand for and Supply of Credit (restrictions not binding)
Restrictions: B(1, 1)=1, B(2, 2)=1, B(1, 5)=0, B(2, 4)=0 Tests of cointegration restrictions: Hypothesized Restricted LR Degrees of No. of CE(s) Log-likehood Statistic Freedom Probability
2 153.1851 NA NA NA 2 Cointegrating Equation(s): Convergence achieved after 1 iterations. Restricted cointegrating coefficients (std.err. in parentheses)
LDCP LR DR LLC LGDP 1.000000 -0.057599 0.110920 -0.803948 0.000000 (0.00000) (0.00439) (0.00833) (0.01084) (0.00000) 2.695212 1.000000 -1.121912 0.000000 -4.837689 (0.24606) (0.00000) (0.06638) (0.00000) (0.23624)
Adjustment coefficients (std.err. in parentheses)
D(LDCP) -2.689352 -0.113378 (0.75330) (0.04559)
D(LR) -11.72357 2.105997 (19.8566) (1.20172)
D(DR) -7.377498 1.077035 (12.6709) (0.76684)
D(LLC) -1.963774 -0.115469 (1.20117) (0.07269)
D(LGDP) -0.853926 0.066335 (0.69096) (0.04182)
LDCP = log banks’ credit to the private sector. LR = lending interest rate. DR = deposit interest rate. LLC = log lending capacity of the commercial banks. LGDP = log gross domestic product.
- 196 -
Table14: Demand for and Supply of Credit (binding restrictions)
The credit supply function is,
)11(8.0113.0059.0433.0 ttttt ullcdrlrldcp ++−+=
and the inverse credit demand function is ,
)12(726.4125.158.2037.11 ttttt elGDPdrldcplr +++−−=
The lending interest rate is the price of the credit quantity, its long run coefficient is the
slope coefficient of the supply function. Thus, in moving along supply curve the credit
Chi-square(1) 0.870511 Probability 0.350814
Cointegrating Eq: CointEq1 CointEq2
LDCP(-1) 1.000000 2.583479 (0.25265) [ 10.2256]
LR(-1) -0.059108 1.000000 (0.00458) [-12.8999]
DR(-1) 0.112590 -1.125129 (0.00875) (0.06875) [ 12.8696] [-16.3646]
LLC(-1) -0.800895 0.000000 (0.01140) [-70.2735]
LGDP(-1) 0.000000 -4.726127 (0.24175) [-19.5493]
C -0.432690 11.03784
Error Correction: D(LDCP) D(LR) D(DR) D(LLC) D(LGDP)
CointEq1 -2.681641 -0.796960 0.000000 -1.309150 -0.681326 (0.78682) (10.6773) (0.00000) (0.44925) (0.68084) [-3.40822] [-0.07464] [ NA ] [-2.91411] [-1.00071]
CointEq2 -0.115688 2.309846 1.232735 -0.109057 0.070887 (0.04931) (1.28146) (0.80438) (0.07639) (0.04533) [-2.34635] [ 1.80251] [ 1.53253] [-1.42762] [ 1.56388]
Diagnosis results: LR Test for the Restrictions χ2(1) = 0.87, Prob. = 0.351 Autocorrelation Test χ2(200) = 275.91, Prob. = 0.000 Serial Correlation Test χ2(25) = 36.58, Prob. = 0.063
- 197 -
quantity supplied will increase as the lending rate increases. Whereas, in moving along the
demand curve, the quantity demanded of the credit falls as the lending rate increases. The
long run coefficients of the two curves above are fully identified. Deposits interest rate,
lending capacity and the nominal GDP are shifting factors of the curves, ceteris paribus. An
increase in the costs of mobilizing savings will increase the deposit interest rate, which in
turn will reduce the banks’ supply of credit to the private sector; while an increase in the
deposit interest rate shifts the demand curve outward. A positive shock to the lending
capacity shifts the supply curve outward, and a positive shock to GDP shifts the demand
curve outward. The quantity of credit supplied to the private sector depends substantially on
the input costs and the lending capacity. The lending rate has the smallest effect on the
credit supply. The lending rate contains small information on how much credit the banks
can supply. That is, the credit supplied is very much influenced by the costs of and how
much savings can be mobilized and formed into the lending capacity. This evidence tallies
well with conclusions made in chapter two, I argued that the banks’ credit to the private
sector is supply-constrained. The banks have problems of mobilizing enough savings to
meet the credit demands of the economy. Repressed interest rates and deliberate
suppression of the banking growth and expansion have limited the growth of the lending
capacity of the banks. The credit quantity does converge towards the equilibrium. If the
credit quantity increases in this period above the equilibrium level, it will fall in the next
period. The shifting variables, deposit interest rate, the lending capacity and the GDP seem
to receive no feedback effects from the system. The two curves, demand and supply, are
identified based on the restrictions made on certain long run coefficients, but the
restrictions are found not binding in table 13, they became binding in table 14 after
restricting that the deposit interest rate is weakly exogenous with respect to the estimation
of the supply function. The deposit interest rate is thus an important explanatory variable
for how much credit the bank can supply. Table 14 requires improvements, there is
autocorrelation problem up to lag four ( the maximum lag used in the diagnosis test) at 1 %.
This problem could result from inadequate modeling and omission of relevant variables.
For example problem loans are relevant factors in the supply determination, and private
sector debt and the business profitability are relevant factors in the demand determination.
Therefore, additional information is required to make the model adequate and reduce the
- 198 -
problems of autocorrelation or serial correlation. We do have a data constraint, we have no
reliable data on the problem loans of the banks, and similarly no data exist for the private
sector debt measures. I use a country risk measure to represent the private sector debt. The
squared government debt, GDS, measures the country’s risk, and it is expected to influence
both the supply and demand functions of the bank credit volume. The vector auto
regression now consists of,
Zt = (LDCPt, LRt, DRt, LLCt, , LGDSt, LGDPt)
The trace co-integration test produces four relations at 5 % significance level and three
relations at 1 % significance level; while, the Max-eigenvalue test produces three relations
at both 1 % and 5 % significance levels, Appendix A16. We cannot identify the third and
fourth relations. We hold the second relation as demand and supply functions and impose
the necessary restrictions for their identification. Under the second relation and given our
restrictions on the long run coefficients, the first co-integration row is treated as the supply
function and the second co-integration row is treated as the demand function, table 15. The
variable, LGDS, under the unrestricted co-integration, appears significant in the supply
function and insignificant in the demand function, and this motivates us to set its long run
coefficient to zero in the demand curve. Its sign however does not conform to our
expectation. It is expected to factor the country’s risk in the model, and this implies that
LGDS should have negative coefficients in both the supply and the demand curves. On the
contrary, the run coefficients of LGDS are significantly positive in both functions. We also
set to zero all those adjustment coefficients that appear insignificant in the unrestricted co-
integration. These restriction results are in the following identified system in table 15:
- 199 -
Table15: Demand for and Supply of Bank Credit (final Model)
The credit supply function is,
)13(006.0735.0113.007.0541.0 tttttt ulGDSllcdrlrldcp +++−+=
and the inverse credit demand function is ,
)14(462.3148.1317.1913.8 ttttt elGDPdrldcplr +++−−=
Cointegrating Eq:
CointEq1 CointEq2
LDCP(-1) 1.000000 1.317149 (0.75221) [ 1.75105]
LR(-1) -0.070316 1.000000 (0.00852) [-8.25216]
DR(-1) 0.113167 -1.148085 (0.01715) (0.15190) [ 6.59914] [-7.55832]
LLC(-1) -0.735276 0.000000 (0.02328) [-31.5852]
LGDS(-1) -0.005706 0.000000 (0.00214) [-2.66449]
LGDP(-1) 0.000000 -3.462451 (0.75111) [-4.60981]
C -0.541409 8.913072
Error Correction: D(LDCP) D(LR) D(DR) D(LLC) D(LGDS) D(LGDP)
CointEq1 -1.093519 7.982792 0.000000 -0.608879 0.000000 -0.679223 (0.33173) (3.02811) (0.00000) (0.27855) (0.00000) (0.25221) [-3.29640] [ 2.63623] [ NA ] [-2.18585] [ NA ] [-2.69308]
CointEq2 -0.090675 1.166470 0.732216 -0.097189 0.000000 0.000000 (0.02764) (0.51308) (0.36391) (0.03046) (0.00000) (0.00000) [-3.28020] [ 2.27349] [ 2.01206] [-3.19070] [ NA ] [ NA ]
LR Test for the Restrictions χ2(5) = 12.825, Prob. = 0.025 Autocorrelation Test χ2(288) = 369.945, Prob. = 0.000 Serial Correlation Test χ2(36) = 35.658, Prob. = 0.485 Heterokedasticity Test χ2(588) = 598.086, Prob. = 0.378
- 200 -
In table 15, the LM test indicates no serial correlation problem, while the Portmanteau test
still rejects the null hypothesis of no serial correlation. It is however better than table 14,
where both tests of Portmanteau and LM tests indicate the existence of serial correlation
problem. The long run coefficients in the system are identified according to the restrictions
we imposed and they are significantly different from zero.
The banks’ credit supply to the private sector in the Gambia is determined by three factors
namely the interest incomes the banks will earn on the loans, the deposit interest incomes
the banks must pay on the deposits and how much demand deposits and capital can be
transformed into the lending resources for the banks. Lending rate has smaller magnitude of
impact on the credit supply, which means the supply curve is very steep. Ceteris paribus,
for the quantity of credit supplied to increase by 7 per cent the lending interest rate must go
up by almost 100 per cent. Thus, lending rate cannot be a good policy for injecting
substantial increases in the banks’ credit supply to the private sector. The deposit interest
rates are costs to the banks, they reduce the banks profits from the loans; and most
importantly they reduce the terms of maturity of the deposits in the banks. Thus, the
increased input costs and the reduced terms of maturity limit the banks’ ability to supply
credit to the private sector. A 10 per cent increase in the deposit interest rates reduces the
credit supply by 1.13 per cent. The most important determinant of the credit supply of the
banks is the lending capacity; ceteris paribus, a 10 per cent increase of the lending capacity
will lead to a 7.35 per cent increase in the supply quantity of the credits. The lending
capacity is found to be an endogenous variable in the system; it will increase by 0.61 per
cent in the next period following a 1 per cent increase in the credit supply. The income
level, GDP, is endogenous in the system. Bank ‘s credit activities feed back to the income;
if the credit supply increases 1 per cent in the current period, the income level may increase
by 0.68 per cent in the subsequent period. The squared government debt does not act as a
risk factor in the supply function, it has a very small impact on the credit supply, but it
exposes a reality in the Gambia over this study period. The agricultural sector has been a
dominant sector, and the produce purchases were financed by loans guaranteed by the
government, while the central bank was a financial arm for the government. This created a
situation where the government expenditures could create their supplies of credits from the
banking sector, and hence government spending has positive association with the bank
- 201 -
credit supply. The supply curve is very inelastic with respect to the lending rate; that is,
other determinants have more influence on the credit supply than the price of credit. The
banks cannot respond sufficiently to changes to the lending rate given that the other
determinants, such as the lending capacity and the deposit interest rate do not change. The
demand for bank credit is determined by the lending rate, the deposit interest rate and the
nominal GDP. Changes in the lending rate induce changes in the quantity of bank loans
demanded, while changes in the other determinants induce changes in the demand curve
itself. The private sector’s demand for bank loans depends highly on the level of economic
activities, a 1 per cent increase in the income will cause the private sector to increase their
demand for bank loans by 2.63 per cent. Also increasing deposit interest rates will
encourage the private sector to spare the savings and demand for bank loans. The private
sector cares more about the returns on the savings than the lending rate with respect to the
demand for bank loans. That is, when they demand for bank loans, they consider first the
level of economic activities and the returns on those activities before the interest rates the
banks charge on the loans. Similarly, it was found that the banks’ credit supply curve is
very steep. Comparing the two curves, we find that the demand curve is flatter than the
supply curve; which means the bank credit as the lending rate is concerned, is supply
constrained. According to our models in chapter 4, the credit policy will be more effective
if the demand curve is flatter than the supply curve. In the elasticity terms, this means the
interest rate elasticity of demand is absolutely larger than the interest rate elasticity of
supply. Next, we examine the assumptions and the hypotheses constructed in chapter 4
about the transaction costs.
Real Money Balance and the Transactions Costs
We maintain in chapter 4 that the entrepreneurs will demand to hold real money balance if
the transactions costs associated with obtaining credit goes up. That is, if the transaction
costs increase the entrepreneurs will hoard real money balances to provide liquidity and
financing for their economic transactions. Thus, in the demand function of the real money
balance, I include the transactions costs. I use the spread between the lending interest rate
and the deposit interest rate to measure the transaction costs as we have defined it in
chapter 4, section 4.4. One co-integration relation exists, Appendix A17, and its long run
- 202 -
coefficients have the correct signs, increased real output, LRGDP, increases the demand for
the real money balances, increased interest rate, LI, reduces the demand for the real money
balances, and increased transactions costs, ID, increase the demand for the real money
balances. The interest rate and the transaction costs are found to be weakly exogenous in
the system, they can explain exogenously changes in the real money balances. Table16
presents the results.
Table16: Real Money Demand Function
Real Money Demand Function,
)15(06.0058.0122.1812.6 ttttt zidlilRGDPlrm ++−+−=
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
LRM(-1) 1.000000
LRGDP(-1) -1.121511 (0.12449) [-9.00868]
LI(-1) 0.058513 (0.01942) [ 3.01274]
ID(-1) -0.060366 (0.03010) [-2.00555]
C 6.812340
Error Correction: D(LRM) D(LRGDP) D(LI) D(ID)
CointEq1 -0.329738 0.305747 0.000000 0.000000 (0.14678) (0.07210) (0.00000) (0.00000) [-2.24644] [ 4.24057] [ NA ] [ NA ]
LRM = log real money supply, M1. LRGDP = log real gross domestic product. LI = lending interest rate; it is the same as LR. ID = spread between LR and DR, the proxy for the transaction costs Diagnosis results: LR Test for Restriction χ2(2) = 2.184, Prob. = 0.335 Autocorrelation Test χ2(80) = 83.176, Prob. = 0.3819 Serial Correlation Test χ2(16) = 23.764, Prob. = 0.095 Normality JB Test χ2(8) =110.64, Prob. = 0.000 Heterokedasticity Test χ2(20) = 20.88, Prob. = 0.404
- 203 -
We examine the irrelevance of the measure of transactions costs in this model, and the
restriction is rejected, and we cannot reject the restriction that the transaction costs are
weakly exogenous with respect to this system. Also, the lending interest rate is weakly
exogenous, while the real income is endogenous. The money demand depends relatively
more on the level of real economic activities than on the interest rate and the transaction
costs. In fact, the evidence shows that money demand is interest inelastic, 1 per cent change
in the interest rate can induce only 0.06 per change in the demand for money; thus the LI is
empirically here a small negative value. This could not be surprising evidence in the case of
the Gambia, where the interest rates may not change properly according to the market
forces, they are instead managed, this might account for the small value of LI.
Transactions Costs, Income and Price level
The credit models in chapter 4 also maintain that increased income, LGDP, reduces the
transactions costs, while increased price level, LCPI, increases the transactions costs. If
these hypotheses cannot be rejected plus the previous assumption that increased
transactions costs increase the demand for real money balances, then the positive impact of
credit on the output, chapter 4, sections 4.4.2 and 4.4.3, will not change. These three
conditions form the foundation blocks of our credit models, if they are rejected, the
predictions of the models that credit as a policy can promote output expansion will be
questionable. The data from the Gambia have proven that the hypotheses cannot be
rejected. Both trace and max Eigenvalue tests indicate the existence of one co-integration
relation for the set of transaction costs, real income and price level, Appendix A18. In the
table below, we find that increased real income reduces the transactions costs, while
increased price level increases the transactions costs. The transactions costs are found to
feed back onto the price level and not onto the income. Table17 presents the unrestricted
results.
- 204 -
Table17: Transaction Costs Function (unrestricted Co-integration)
The income is found to be weakly exogenous with respect to the system. Income level
explains to a great degree the level of transactions costs in the economy, a 1 per cent
increase in the income leads to 4.5 per cent fall in the transactions costs. The price level is
not weakly exogenous, and it account for significant changes in the transactions costs; 1 per
cent increase in the price level leads to 4.46 per cent increase in the transactions costs. The
two long run coefficients are very much close raising the question that the two variables
could be offsetting each other. The increase in the income that could lead to reduced
transaction costs may be offset by an equivalent increase in the price level if the price level
is allowed to change. The test results are found below:
1 Cointegrating Equation(s): Log likelihood 41.44228
Normalized cointegrating coefficients (std.err. in parentheses) ID LRGDP LCPI
1.000000 4.492106 -4.465253 (0.82459) (0.28713)
Adjustment coefficients (std.err. in parentheses) D(ID) -0.150859
(0.13193) D(LRGDP) -0.008920
(0.00853) D(LCPI) 0.034433
(0.00481)
ID = spread between LR and DR, the proxy for the transaction costs. LRGDP = log real gross domestic product. LCPI = log consumer price index.
- 205 -
Table18: Transaction Costs Function ( restricted co-integration)
Transaction Cost Function,
)16(437.4387.428.29 tttt elcpilRGDPid ++−=
Increased prices obscure the information and increase the uncertainty in the economy, and
hence increase the costs of the economic transactions. On the other hand, increased
transaction costs also cause the prices to rise. The only control variable in the model is the
real income. If the real income can rise high and above any rise in the price level, the
economic agents will be better off because the costs of the economic transactions could
remain unchanged or decline.
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
ID(-1) 1.000000
LRGDP(-1) 4.386593 (0.82514) [ 5.31616]
LCPI(-1) -4.437308 (0.28733) [-15.4435]
C -29.27752
Error Correction: D(ID) D(LRGDP) D(LCPI)
CointEq1 -0.177528 0.000000 0.034440 (0.13035) (0.00000) (0.00484) [-1.36193] [ NA ] [ 7.11254]
Diagnosis results: LR Test for Restriction χ2(1) = 1.123, Prob. = 0.289 Autocorrelation Test χ2(36) = 59.81, Prob. = 0.007 Serial Correlation Test χ2(9) = 10.015, Prob. = 0.349 Normality JB Test χ2(6) = 116.014, Prob. = 0.000 Heterokedasticity Test χ2(36) = 46.74, Prob. = 0.108
- 206 -
5.3.8 Schumpeter’s Credit
Schumpeter’s credit is the credit the banks provide to the entrepreneurs to enable them to
undertake new production activities not those already in the economic circulation. It is
created in an anticipation of an economic transaction. Thus, the increase of this credit
should feed back to the entrepreneurial investments. We run a co-integration test for the
private investment, a proxy for the entrepreneurial investments, and the banking variables
of bank credit to the private sector, the lending capacity and the lending interest rate. That
is, the entrepreneurial investments are expected to depend on the banks’ credit, the lending
capacity of the banks and the interest rate. The co-integration tests, trace and Max-
eigenvalue, indicate one co-integration relation in the system, Appendix A19. This relation
is normalized by the long run coefficient of the private investments as in Table19.
Table19: Schumpeter’s Credit Function (unrestricted co-integration)
The long run coefficient of lending capacity is not significant, while its adjustment
coefficient is significant indicating that the private investments feed back to the lending
capacity of the banks. Banks’ lending capacity will be boosted following an increase in the
private investments. Also, private investments feeds back to the banks’ credit; banks’ credit
to the private sector significantly reacts to the private investments. It implies that bankers
1 Cointegrating Equation(s): Log likelihood -40.25640
Normalized cointegrating coefficients (std.err. in parentheses) LRPI LRDCP LRLC LI
1.000000 -1.198366 0.395153 -0.055560 (0.32733) (0.24724) (0.01005)
Adjustment coefficients (std.err. in parentheses) D(LRPI) -0.182319
(0.10612) D(LRDCP) 0.447028
(0.08773) D(LRLC) 0.326327
(0.07944) D(LI) 0.144441
(1.62713)
LRPI = log real private investments. LRDCP = log real banks’ credit to the private sector. LRLC = log real lending capacity of the commercial banks. LI = lending interest rate.
- 207 -
do not anticipate private investments, they lag behind the entrepreneurs, and hence they do
not supply Schumpeter’s credit. In fact, the restrictions that banks’ credit is weakly
exogenous is rejected, only the lending interest rate is found to be weakly exogenous with
respect to the system as table 20 illustrates.
Table 20: Schumpeter’s Credit Function (restricted co-integration)
Private Investment Function,
)17(044.0668.0868.0 tttt ulilrdcplrpi +++=
The positive coefficient of the lending rate in this function cannot be accepted theoretically.
The private investment cannot be an increasing function of the lending rate, but this is what
the evidence here shows. This explains that the lending interest rate of the banks is
irrelevant to the amount of the investments the entrepreneurs undertake, it is the quantity of
credit that concerns the private sector. They are willing to pay for the increased lending
interest rate as long as the banks can deliver the required credit quantity; the entrepreneurs
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq: CointEq1
LRPI(-1) 1.000000
LRDCP(-1) -0.668573 (0.11209) [-5.96434]
LRLC(-1) 0.000000
LI(-1) -0.043957 (0.00632) [-6.95697]
C -0.868227
Error Correction: D(LRPI) D(LRDCP) D(LRLC) D(LI)
CointEq1 -0.290492 0.468195 0.372741 0.000000 (0.11897) (0.10949) (0.09299) (0.00000) [-2.44178] [ 4.27626] [ 4.00833] [ NA ]
LR Test for the Restrictions χ2(2) = 0.796, Prob. = 0.672 Autocorrelation Test χ2(80) = 77.88, Prob. = 0.546 Serial Correlation Test χ2(16) = 14.41, Prob. = 0.568 Heterokedasticity Test χ2(20) = 22.967, Prob. = 0.290
- 208 -
will incorporate the lending interest rate into their prices of the investment outputs. The
other explanation for the positive association between the private investments and the
lending interest rate could be that the lending interest rate has been for a long time either
fixed or managed, and thus it does not reflect the market interest rate; the positive sign is
hence spurious. It is evidence that the banks have lost connection with real activities in the
private sector; they do not lead the private sector activities, they instead follow the
developments. The banks’ credit quantity to the private sector does not also have an elastic
impact on the private investments, 1 per cent increase in the banks’ credit to the private
sector will induce less than 1 per cent increase in the private investments. This implies a
credit constraint exists in the Gambia, but it is a weak constraint. The credit quantity of the
banks carries better information in the long run for the private sector than the lending
interest rate.
5.3.9 Monetary Policy Competence
A negative one-to-one relationship between domestic credit and foreign assets indicates
that the domestic monetary policy is incompetent, because a change in the country’s credit
holdings will be completely offset by an equal change in the foreign assets. I run a bilateral
co-integration test between the domestic credit and the foreign assets, and two co-
integration relations are found at 5% significance level and one co-integration relation at
1% significance level, Appendix A20. The two variables are found to be I(1) processes,
table 1. Thus, we consider only the co-integrating relation at 1% significance level. Table
21 below gives the unrestricted relation, and table 22 gives the restricted one:
- 209 -
Table 21: Domestic Credit versus Foreign assets (unrestricted co-integration)
For a negative one-to-one relation to hold between the domestic credit and the foreign
assets in the long run, then the summation of their long run coefficients must equal zero.
This restriction result is in table 22. The hypothesis is rejected at 1% significance level; and
thus there is no negative one-to-one relation between the foreign assets and the domestic
credit of the Gambia. This indicates that the monetary policies of the Central Bank over this
sample period were effective and competent. Taking other variables into account, the
monetary policy is reactionary with respect to changes to the exchange rate and the foreign
assets, it does not anticipate changes in these variables, the monetary authorities act after
the fact. The foreign assets here constitute those held by the commercial banks and the
central bank, and thus the negative sign between the domestic credit and the foreign assets,
table 21, could imply that the monetary authorities care more about the interest rates than
about the foreign reserves of the Central Bank (Roubini, 1988).
1 Cointegrating Equation(s): Log likelihood -242.8418
Normalized cointegrating coefficients (std.err. in parentheses) CDC CNFA
1.000000 0.660513 (0.10022)
Adjustment coefficients (std.err. in parentheses) D(CDC) 0.075983
(0.09878) D(CNFA) -0.380996
(0.06388)
CDC = domestic credit as a percentage of GDP. CNFA = net foreign assets as a percentage of GDP
- 210 -
Table 22: Domestic Credit Versus Foreign Assets (restricted co-integration)
The research findings are summarized in chapter 6. Chapter 6 gives the concluding points
of the research, make the recommendations and identifies the future research directions.
Tests of cointegration restrictions:
Hypothesized Restricted LR Degrees of No. of CE(s) Log-likehood Statistic Freedom Probability
1 -252.4169 19.15026 1 0.000012
1 Cointegrating Equation(s): Convergence achieved after 1 iterations.
Restricted cointegrating coefficients (not all coefficients are identified) CDC CNFA
0.005906 -0.005906
Adjustment coefficients (std.err. in parentheses) D(CDC) -12.76295
(5.63947) D(CNFA) -4.067609
(5.40875)
Autocorrelation Test χ2(16) = 9.477, Prob. = 0.893 Serial Correlation Test χ2(4) = 0.94, Prob. = 0.92 Heterokedasticity Test χ2(12) = 30.54, Prob. = 0.0024
- 211 -
6 Conclusions and Recommendations
This chapter gives a summary of the main results from chapter 5. Chapter 5 has discussed
the survey and the empirical findings with respect to the theoretical models of chapter 4.
The conclusion of chapter 2 on the background review has resolved our first research
problem, section 1.1.1, and we need not repeat it here. The second research problem,
section 1.1.2 is discussed in the survey findings, section 5.2, and we give its main results in
this chapter, and the empirical findings with respect to the final research problem, section
1.1.3 is discussed in section 5.3, and we summarize the main conclusions here. We then
derive the recommendations and the future directions of the research.
The survey indicates that the banks tend to finance consumption goods more than the
capital goods. The public are not attracted by the interest rate to save in the banks; thus,
they seldom hold time deposit accounts. They hold current accounts in order to make easy
payments through the banking system. The current state of the banking services do not
impress the public; banks are not reaching to the wider population of the country.
The empirical results show that the banking activities have no long run relationships with
either the private investments or the gross domestic investments, Appendix A4 and
Appendix A5. This confirms the evidence from the survey that banking industry tends not
to engage in the investment financing. There are long run relationships between the banking
activities and the output, section 5.3.2, Appendix A1, between the banking activities and
the capital accumulation, section 5.3.3, Appendix A3, and between the banking activities
and the economic efficiency, section 5.3.4, Appendix A6. In these relationships banking
activities are found to promote output and capital accumulation but decrease the economic
efficiency. In section 5.3.5, we allow for control variables so that any economic variable
could co-integrate not only with the set of the banking variables but also with the set of
other economic variables. In this system, we find that an increase in the banks’ credit to the
private sector is associated with increases in the output, the physical capital accumulation
and the economic efficiency. The liquid liabilities are also found to increase the output;
while the banks’ assets and the liquid liabilities reduce both the physical capital and the
economic efficiency in the long run, table 11. The set of banking variables is found not
weakly exogenous, it significantly reacts to the disequilibrium in the system, Appendix A9.
- 212 -
We however find that the causality runs from the banking to the economic variables. The
log real per capita of banks’ credit to the private sector and the log real per capita of liquid
liabilities are found to be Granger causal for both the log real per capita output and the log
real per capita economic efficiency, and no other variables in the system is found to be
Granger causal for them. Thus, these two banking variables account for significant changes
in the real economic activities. Yet, the banks’ credit to the private sector does not qualify
to be a Schumpeter’s credit that starts the circular flow of creditary production and enables
the entrepreneurs to carry out new production activities. This credit is expected to be a
significant explanatory variable in the investment function of the entrepreneurs as defined
in section 4.2, we estimate this function and table 20 presents the results. There is no
evidence that the banks in the Gambia supply the Schumpeter’s credit, the banks’ credit to
the private sector is found to be not weakly exogenous. A weak credit constraint as defined
in section 4.4.3 also exists, the long run coefficient of the banks’ credit quantity is
significantly smaller than one. The constraint is due to the inability of the banks to change
the credit supply as the interest rate changes. The credit supply is found to be interest rate
inelastic, and the lending capacity and the deposit interest rate have relatively larger
influences on the credit quantity than the lending interest rate, table 14 and table 15. The
weak credit constraint also implies the private sector may accept an increased bank credit
supply only with a reduced interest rate, Chapter 4, section 4.4.3. On the other hand, the
demand for the bank credit is largely determined by the level of the economic activity. The
bank credit market is found to be grossly inefficient, the spread between the lending interest
rate and the deposit interest rate, which in this research measures for the transaction costs,
tends to grow large. It is found to increase the inflation, table 18, and causes the public to
increase their demand for real money balances, table 17. We find that the banking industry
is far off the competitive and efficient path; the empirical reserve ratio of 67% is
substantially larger than the mean regulatory reserve ratio of 19.27% over the period 1964 –
2002. In a perfectly competitive banking industry, the empirical reserve ratio will equal the
regulatory reserve ratio. Perfect competitive banking industry may not exist, but it is a
benchmark for evaluation and comparison. Thus, the larger the deviation of the empirical
reserve ratio from the regulatory reserve ratio, the less competitive and the less efficient
becomes the banking industry. This deviation in the case of the Gambia amounts to 250%
- 213 -
of the regulatory reserve ratio, and hence the banking industry is far off the competitive
path. Yet, its credit to the private sector is vital for the economy, it increases and causes the
output expansion and improves the efficiency in the long run; it tends to increase the
physical capital but not Granger cause it. For the central bank, we find that it has faced
difficulties due to its conflicting objectives of simultaneously pursuing price stability,
increased employment and supporting the government finance, chapter 2, figure 2.1; but the
competence of its monetary policy over the study period cannot be rejected, table 22. From
these findings and conclusions the research can derive some recommendations for the
authorities.
The declining trend of banks’ credit to the private sector is worrisome, since any change in
this variable could have a permanent impact on the economic output, the capital stock and
the efficiency. The banks should reverse this declining trend, they should reduce the
consumption loans and their holdings of the central bank bills and the government treasury
bills. The central bank and the commercial banks should work together to reduce the spread
between the lending interest rate and the deposit interest rate. This spread reduces the
savings in the banking system, encourages the public to hold increasingly high amounts of
cash money and increases the inflation. The increased spread also indicates the inability of
the banks to search and select profitable projects and monitor their credit portfolios. The
banks should increase and improve their personnel to be able to supply credits that depend
on the expected profits of the private sector investments not on collaterals and relationships.
In the history of the Gambia banking industry, few banks employed just over hundred
people, and thus the nature of their inputs of low human resources and low savings reflects
to a great extent the nature of their output of low credit quantities to the private sector with
high collateral and securities.
The banks have to learn and adopt the norms and values of the society they serve. Their
schemes of savings mobilizations or deposit collections should be inclusive and
comprehensive not exclusive and selective as some banks currently practice. The banks
should learn that interest rate is not a prime objective of the society for placing money in
the banks. The society is concerned first with payment system of the banking industry
followed by the safety of money in the banks and then the savings for the future. The
interest rate is the least reason for placing money in the banks. The banks’ provision of the
- 214 -
means of payment is vital for the economic development; and it reduces the risk of theft
and corruption associated with carrying cash around to make payments. The society wants
their money and savings to be safe in the banks, that they can have their claims back any
time they wish. Thus, the banking system should be nationwide; every member of the
society should feel the presence of the banks, every member is either to receive bank
money, for example bank credit, bank cheques, guarantees and acceptances, or he is to pay
in bank money. The trade seasons of the agricultural sector should be ‘banknized’, the
farmers should be paid in banking means of payment that could be converted immediately
into cash money any time they wish. The society also wants to save for the future. For the
banks to mobilize these savings, they should provide correspondingly long term goods for
the society.
The banking industry should embark on the universal banking system to build long term
commitments with the society. Also, they should allow the society to share in the profits
and dividends of the banks. This could be achieved by setting up a stock market in the
country where the shares of the banks could be floated and the ordinary public could be
allotted large proportions . The economic development could be further spurred and the
wealth of the nation would expand and tend to be distributed evenly if the central bank
could have a branch in every administrative division to help the banking services and
output get close to the people instead of concentrating the banking industry and
consequently the wealth in one area and then as a result implicitly forcing the people to
move and migrate close to that particular area.
The banks should do more beyond the standard operating procedures and the existing
circular flow. They should create Schumpeter’s credit that creates entrepreneurs and
enables them to carry out new economic activities. The banks should promote small
enterprises that employ large amounts of labor instead of concentrating on financing capital
accumulation that could harm the economy in the long run, particularly in the Gambia
where a good percentage of the population is illiterate. The goal of making the Gambia a
financial center depends not only on having a competitive and efficient banking industry
but also on having a large percentage of the population owning to the banks. People will
own to the banks if the small enterprises have access to the banking facilities. Our economy
is in the early stages of economic development and for the banks to gear up the economic
- 215 -
development and prosperity and be the engines of growth, they should do what banks did in
the early stages of today’s developed economies. That is, they should be promoters and
financiers of the enterprises.
This research is intended to constitute some sections in a book the researcher plans to write
on the structures of the Gambian economy. It will be further developed to include the
microeconomic analysis of the bank performances and to characterize the projects and the
borrowers that qualify for the bank credits; then, we will sample and appraise the past loan
decisions of the banks. Broaddus’ model in section 4.3 will be estimated, if data are
available, for all the sub-Saharan African countries for the comparison analysis. The model
in section 4.4.3 that interlinks credit and money will be estimated and calibrated in a small
open economy. The interactive relationships between the fiscal policy and the monetary
policy in sections 4.4.1 and 4.4.2 will be empirically investigated. By incorporating these
developments, a major part of the book, the structures of the Gambian economy, on the
financial system will be completed.
- 216 -
References
1. Allen, Franklin, and Santomero, Anthony M.; 1993, „What Do Financial
Intermediaries Do?”, the Warton School, University of Pennsylvania, Working
Paper (99-30-B).
2. Barro, Robert J.; 1978, “Unanticipated Money, Output, and the Price Level in the
United States.” Journal of Political Economy 86(4), (August), pp. 549 – 580.
3. Barro, Robert J., and Sala-I-Martin, Xavier; 1995, “Economic Growth”, McGraw-
Hill, Inc.
4. Beddies, Christian H.; 1999, “Investment, Capital Accumulation, and Growth:
Some Evidence from the Gambia 1964 – 98”, IMF working paper, 99/ 117,
Washington DC.
5. Benassy, Jean-Pascal; 2002, “The Macroeconomics of Imperfect Competition and
Nonclearing Markets: A Dynamic General Equilibrium Approach”, Massachusetts
Institute of Technology.
6. Bencivenga, V. R. and Smith, B. D.; 1991, “Financial Intermediation and
Endogenous Growth”. Review of Economic Studies 58(April) s. 194, pp. 195 – 209.
7. Bernanke, Ben S. and Blinder, Alan S.; 1988, “Credit, Money and Aggregate
Demand“, Working paper No. 2534, national Bureau of Economic Research, Inc.
8. Bolton, Patrick and Dewatripont, Mathias; 2005, “Contract theory”, Massachusetts
Institute of Technology.
9. Boschini, Anne D.; Pettersson, Jan and Roine, Jesper; 2003:12, “Resource Curse or
Not: A Question of Appropriability”, Working Paper , Department of Economics,
Stockholm University.
10. Bossone, Biagio, and Sarr, Abdurrahamane; 2002, “A New Financial System for
Poverty Reduction and Growth”, IMF WP/02/178.
11. Broaddus, Alfred; October 16, 1973, “Monetary Policy, Banking Credit, and Total
Credit: A Preliminary Analysis”, Working Paper 73-2, Federal Reserve Bank of
Richmond.
12. Brunner, Karl and Meltzer, Allan H.; 1988, “Money and Credit in the Monetary
Transmission Process”, American Economic Review Vol. 78, s.2.
- 217 -
13. Cameron, Rondo; Crisp, Olga; Patrick, Hugh T. and Tilly, Richard; 1967, “Banking
in the Early Stages of Industrialization”, Oxford University Press, Inc.
14. Carmichael, Jeffery; 1982. “Money and Growth: Some Old Theories from a New
Perspective”. Economic Record 58(163) (December), pp. 386 – 394.
15. Catao, Luis; 1997. “Bank Credit in Argentina in the Aftermath of the Mexican
Crisis: Supply or Demand Constrained?”, IMF WP/97/32.
16. Chiang, Yeong-Yuh, November 1994 “ Information, Efficiency and Intermediation
“ Discussion Paper no. 9418, the Institute of Economics Academia, Sinica,
Nankang, Taiwan, Republic of China.
17. Cumby and Obstfeld, Maurice; September 1981, “Capital Mobility and the Scope
for Sterilization: Mexico in the 1970s”, Working paper No. 770, National Bureau of
Economic Research, Cambridge, MA 02138.
18. Easterly, Williams and Levine, Ross; 1997, “Africa’s Growth Tragedy: Policies and
Ethnic Division”, Journal of Economics 112 (November) 1203-50
19. Englmann, F. C.; 1994. “Schumpeterian Model of Endogenous Innovation and
Growth”. Journal of Evolutionary Economics, Bd.4,3, s. 227 – 241.
20. Erhardt, Wolfram; 2002, “Financial Markets for Small Enterprises in Urban and
Rural Thailand: Empirical Analysis on the Demand for and Supply of Financial
Services with Particular Emphasis on the Determinants of Credit Access and
Borrower Transaction costs”. Frankfurt am Main .
21. Esanov, Akram; Buiter, Willem H. And Raiser Martin; 2004 “ Nature’s Blessing or
Nature’s Curse: The Political Economy of Transition in Resource-Based economies
“, Zentrum fuer Entwicklungsforschang discussion paper on development policy,
81. ZEF, Bonn.
22. Fama, Eugene, F.; 1980. “Banking in the Theory of Finance”, Journal of Monetary
Economics 6, 39 – 57.
23. Favero, Carlo A.; 2001, “Applied Macroeconometrics”, Oxford University Press.
24. Felderer, Bernhard and Homburg, Stefan; 1992, “Macroeconomics and New
Macroeconomics”, Spriger-Verlag Berlin Heidelberg, Germany.
25. Frankel, Jeffrey A.; December 1994, “Sterilization of money Inflows: Difficult
(Calvo) or Easy (Reisen)?”, working paper wp/94/159, IMF.
- 218 -
26. Fry, Maxwell. J.; 1995, “Money, Interest and Banking in Economic Development”,
the John Hopkins Studies in Development, Baltimore, John Hopkins University
Press.
27. Fry, Richard; 1975, “Bankers in West Africa, The Story of the Bank of British West
Africa Ltd”, Hutchinson, Benham, London.
28. Galetovic, Alexander; March 1994 “ Financial Intermediation, Resource Allocation
and Long Run Growth “, Princeton University and centro de Economia Aplica
discussion paper Nr. 170.
29. Gambia: Report for the years / Colonial Office, London HMSO, 1890 – 1953.
Period 1904 – 1930, 1954/ 55(1957) – 1962/ 63.
30. Graf, Michael and Karmann, Alexander; January 2001, ”Does Financial Activity
Cause Economic Growth?“, Dresden Discussion Paper Series in Economics No.
1/01, Technische Universitat Dresden.
31. Gujarati, Damodar N.; 1995 “Basic Econometrics “, McGraw-Hill Inc.
32. Gurley, John G., and Shaw, Edward S.; 1960, “ Money in a Theory of Finance”,
The Brookings Institution, Washington, D.C.
33. Harrison, Paul; Sussman, Oren and Zeira, Joseph; 1999.,”Finance and Growth:
Theory and New Evidence.”, Finance and Economic Discussion Series, Division of
Research and Statistics and Monetary Affairs, Federal Reserve Board, Washington
D.C., 1999 – 35.
34. Hausmann, Ricardo and Rigobon, Roberto; December 2002, “An Alternative
Interpretation of Resource Curse: Theory and Implications”, National Bureau of
Economic Research, 1050 Massachusetts Avenue, Cambridge, MA 02138.
35. Haveman, Robert H. and Bershadker, Andrew; 2003, “Human Capital in the United
States from 1975 to 2000: Patterns of Growth and Utilization” W. E. Upjohn
Institute for Employment Research, Kalamazoo, Michigan.
36. Jayaratne, J. and Strahan, P. E.; 1996. “The Finance – Growth Nexus: Evidence
from Bank Branch Deregulation”. Quarterly Journal of Economics 111,
s.3(August), pp. 639 – 670.
37. Jeong, Byeongju; November 1998, “Measurement of Human Capital Inputs Across
Countries: New Method and Results”, Center for Economic Research and Graduate
- 219 -
Education, Charles University, Economics Institute, Academy of Sciences of the
Czech Republic, Working paper series (ISSN 1211-3298), Prague.
38. Johansen, Soren; 1995, “Likelihood-Based Inference in Cointegration Vector Auto-
Regressive Models”, Oxford University Press.
39. King, Robert G. and Levine, Ross; 1993, “Finance and Growth: Schumpeter might
be right”, Quarterly Journal of Economics, 108(August).
40. King, Robert. G. and Levine, Ross; 1993b, “Finance, Entrepreneurship, and
Growth: Theory and Evidence”, Journal of Monetary Economics 32, pp. 513 – 542.
41. Kletzer, Kenneth and Spiegel, Mark; June 1999, “Sterilization Costs and Exchange
Rate Targeting”. Working paper No. PB99-03; Center for Pacific Basin Monetary
and Economic Studies, Economic Research Department, Federal Reserve bank of
San Francisco.
42. Korteweg, Pieter and Van Loo, Peter D.; 1977, “The Market for Money and the
Market for Credit: Theory, Evidence and Implications for Dutch Monetary Policy”,
H. E. Stenfert Kroese B.V., Leiden.
43. Kwan, Simon H. and Eisenbeis, Robert A.; 1996, “An Analysis of Inefficiencies in
Banking: A Stochastic Cost Frontier Approach”. Economic Review: an Annual
Report of the Economic Research Department – San Francisco, California Bank, S.
16 – 26.
44. Leahy, Michael; Schich, Sebastain; Wehinger, Gerf; Pelgrin, Florian and
Thorgeirsson, Thorstein; January 26, 2001 “Contributions of Financial Systems to
Growth in OECD Countries, OECD Economics Department WP No. 280.
45. Leibenstein, Harvey; 1996, “Allocative Efficiency versus ‘X – efficiency”.
American Economic Review 56, S.3 (June), pp. 392 – 415.
46. Levine, Ross and Zervos Sara; 1998, “Stock Markets, Banks and Economic
Growth”, American Economic Review 88 (June) 537-58.
47. Levine, Ross; 1997, “Financial Development and Economic Growth”, Journal of
Economic Literature 35 (June) 688-726.
48. Levine, Ross; 1998, “The Legal Environment, Banks and Long-Run Economic
Growth”, Journal of Money, Credit and Banking, 30(2), 596-620 August.
- 220 -
49. Levine, Ross; Norman, Loayza and Thorsten, Beck; August 2000, “Financial
Intermediation: Causality and Causes”, Journal of Monetary Economics, V.46 No.1,
31-77.
50. Lucas, Robert E. Jr.; 1988, “On the Mechanics of Economic Development”. Journal
of Monetary Economics, Vol.22 (July), pp. 3- 42.
51. Luetkepohl, Helmut; 1991, “Introduction to Multiple Time Series Analysis”,
Springer-Verlag-Berlin Heidelberg.
52. Madonia, Guiseppina; October 1994, ”Testing for Sterilization Policies in Germany,
Italy and UK”. Studies in Economics No. 94/9, Department of Economic,
University of Kent at Canterbury.
53. Manzano, Osmel and Rigobon Roberto; July 2001, “Resource Curse or Debt
Overhang?” Working Paper 8390, National Bureau of Economic Research, 1050
Massachusetts Avenue, Cambridge, MA 02138.
54. Modigliani, Franco, and Miller, Merton H.; 1958, “The Cost of Capital, Corporation
Finance, and the Theory of Investments”, American Economic Review 48, June
261-297.
55. Pissarides, Francesca; November 1998, “Is Lack of Funds the Main Obstacle to
Growth?”., EBRD working paper 33.
56. Report to the Government of the Gambia on Co-operative Banking, Expanded
Programme of Technical Assistance, International Labour Office,
ILO/TAP/Gambia/R.2, Geneva 1964.
57. Romer, Paul; 1993, ”Idea Gaps and Object Gaps in Economic Development”,
Journal of Monetary Economics 32(3) 543-73.
58. Roubini, Nouriel; November 1988, “Offset and Sterilization under Fixed Exchange
Rates with an Optimizing Central Bank”. Working paper No. 2777, National
Bureau of Economic Research, Inc., Cambridge, MA 02138.
59. Rousseau, Peter L. and Wachtel, Paul; November 1998, “Financial Intermediation
and Economic Performance: Historic Evidence from Industrialized Countries”.
Journal of Money, Credit, and Banking, Vol. 30, N. 4, pp. 657-678.
- 221 -
60. Ruhle, Ilonka; 1994, “Why Banks? Microeconomic Foundations of Financial
Intermediaries”, Frankfurt University Dissertation, Peter Lang GHBH, Europaischer
Verlag der Wissenschaften, Frankfurt am Main 1997.
61. Sandberg, Lars G.; September 1978, “Banking and Economic Growth in Sweden
before World War I”, Journal of Economic History 38, 650-680.
62. Schuler, Kurt A.; 1992, “Currency Boards”, Dissertation, George Mason University
Fairfax, Virginia.
63. Schumpeter, J. A., 1989 (English Translation) “Economic Development: An Inquiry
into Profits, Capital, Credit, Interest and the Business Cycles”, Transaction
Publishers.
64. Schumpeter, Joseph Alois; 1934, “The Theory of Economic Development: An
Inquiry into Profits, Capital, Credit, Interest and the Business Cycle”. Translated
from German to English by Redvers Opie. Harvard Economic Series 46.
Cambridge, Massachusetts University press.
65. Shewhart, Walter A. and Wilks, Samuel; 2004, “Applied Econometric Time
Series”, John Wiley & Sons, Inc.
66. Straus, Hubert; 2004, “ Multivariate Co-integration Analysis of Aggregate Exports:
Empirical Evidence for the United States, Canada and Germany”, Kiel Institute for
World Economics, 24100 Kiel.
67. Sylla, Richard; 2000, “Financial System and Economic Modernization: A New
Historical Perspectives”, World Bank Summer Workshop.
68. Tao, Hung-lin, and Stinson, Thomas F.; April 1997, “An Alternative Measure of the
Human Capital Stock”, Bulletin No. 97-1, Economic Development Center,
Department of Economic, Minneapolis, Department of Applied Economics, St.
Paul, University of Minnesota.
69. Tsuru, Kotaru; January 26, 2000, “Finance and Growth: Some Theoretical
Considerations and Review of the Empirical Literature.”, OECD, Economics
Department Working papers, No. 228.
70. Tuuli, Koivu; 2002, “ Do Efficient Banking Sector Accelerate Economic Growth in
Transition Economies”, BOFIT Discussion Papers No.14
- 222 -
71. Walters, Alan and Hanke, Steve H.; April 1992 “ Currency Boards “, Working
Papers in Economics No. 281, the Johns Hopkins university, Baltimore, Maryland
21218.
72. Welsch, Heinz; 2004, “Natural-resource Intensity and Economic Growth in a Model
with Endogenous Technological Change: Theory and Evidence”.
Wirtschaftswissenschaftliche Diskussionspapier, 255; Literaturverz, S. 24 – 25.
73. Winkler, Adalbert; February 1998 “ Financial development, economic growth and
Corporate Governance “, Johann Wolfgang Goethe-Universitaet, Frankfurt, AM
Main, Fachbeieich, Wirtschaftswissenschaften Working Paper series: Finance No.
12.
74. Wohltmann, Hans-Werner; 2005, “Grundzuege der Makrooekonomischen Theorie:
Totalanalyse Geschlossener und Offener Volkswirtschaften”, Oldenburg, Germany.
75. Wray, L.Randall; edited, 2004, “Credit and State Theories of Money: The
Contributions of A. Mitchel Innes”, Edward Elgar Publishing Ltd.
76. Yule, G. Udny and Kendall, M. G. 1950; “ An Introduction to the Theory of
Statistics”, Charles Griffin and Co. Ltd 42 Drury Lane, London WC2.
77. Zahid, Hasan Khan; 1985, “ Monetary Accommodation of Government Budget
Deficits, and Sterilization Analysis policies: An Empirical Analysis of Nine OECD
Countries” PhD Thesis, Columbia University.
- 223 -
Appendix A: Lag Order Selection, Co-integration Rank tests and some Granger Causality tests
A 1: Output and Banking
VAR Lag Order Selection Criteria Endogenous variables: LRY LRPDCP LRPLL LRPCBA Included observations: 35
Lag LogL LR FPE AIC SC HQ
0 5.077493 NA 1.11E-05 -0.061571 0.116183 -0.000210 1 89.69996 145.0671 2.21E-07 -3.982855 -3.094084* -3.676051* 2 107.4632 26.39104* 2.08E-07* -4.083609* -2.483822 -3.531363 3 122.3596 18.72691 2.45E-07 -4.020546 -1.709744 -3.222858 4 136.6568 14.70573 3.32E-07 -3.923245 -0.901426 -2.880114
Series: LRY LRPDCP LRPLL LRPCBA Lags interval (in first differences): 1 to 1
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.575360 52.53511 47.21 54.46 At most 1 0.320375 20.84408 29.68 35.65 At most 2 0.129001 6.554151 15.41 20.04 At most 3 0.038273 1.443915 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at the 5% level Trace test indicates no cointegration at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.575360 31.69103 27.07 32.24 At most 1 0.320375 14.28993 20.97 25.52 At most 2 0.129001 5.110236 14.07 18.63 At most 3 0.038273 1.443915 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 1 cointegrating equation(s) at the 5% level Max-eigenvalue test indicates no cointegration at the 1% level
- 224 -
A 2: Granger Causality, Output and Banking,
VEC Pairwise Granger Causality/Block Exogeneity Wald Tests
Dependent variable: D(LRY)
Exclude Chi-sq df Prob.
D(LRPDCP) 6.172420 1 0.0130 D(LRPLL) 0.123735 1 0.7250
D(LRPCBA) 0.035706 1 0.8501
All 14.98664 3 0.0018
Dependent variable: D(LRPDCP)
Exclude Chi-sq df Prob.
D(LRY) 2.603316 1 0.1066 D(LRPLL) 0.363987 1 0.5463
D(LRPCBA) 0.007434 1 0.9313
All 3.518298 3 0.3184
Dependent variable: D(LRPLL)
Exclude Chi-sq df Prob.
D(LRY) 4.116117 1 0.0425 D(LRPDCP) 1.341135 1 0.2468 D(LRPCBA) 0.207032 1 0.6491
All 6.295531 3 0.0981
Dependent variable: D(LRPCBA)
Exclude Chi-sq df Prob.
D(LRY) 5.184815 1 0.0228 D(LRPDCP) 0.298594 1 0.5848 D(LRPLL) 2.028542 1 0.1544
All 7.582400 3 0.0555
- 225 -
A 3: Capital Stock and Banking
VAR Lag Order Selection Criteria Endogenous variables: LRPK LRPDCP LRPLL LRPCBA
Lag LogL LR FPE AIC SC HQ
0 8.267953 NA 9.21E-06 -0.243883 -0.066129 -0.182522 1 94.69144 148.1545* 1.66E-07* -4.268082* -3.379312* -3.961279* 2 107.5582 19.11634 2.06E-07 -4.089040 -2.489254 -3.536794 3 120.8693 16.73390 2.67E-07 -3.935387 -1.624584 -3.137698 4 140.0588 19.73785 2.74E-07 -4.117648 -1.095829 -3.074516
Series: LRPK LRPDCP LRPLL LRPCBA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.558006 55.64172 47.21 54.46 At most 1 0.387041 24.61624 29.68 35.65 At most 2 0.103332 6.016860 15.41 20.04 At most 3 0.048075 1.872219 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.558006 31.02548 27.07 32.24 At most 1 0.387041 18.59938 20.97 25.52 At most 2 0.103332 4.144641 14.07 18.63 At most 3 0.048075 1.872219 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 1 cointegrating equation(s) at the 5% level Max-eigenvalue test indicates no cointegration at the 1% level
- 226 -
A 4: Private Investment and Banking
VAR Lag Order Selection Criteria Endogenous variables: LRPPI LRPDCP LRPLL LRPCBA
Lag LogL LR FPE AIC SC HQ
0 -14.92392 NA 3.47E-05 1.081367 1.259121 1.142728 1 68.71866 143.3873* 7.32E-07* -2.783923* -1.895153* -2.477120* 2 78.56894 14.63471 1.08E-06 -2.432511 -0.832725 -1.880265 3 87.05792 10.67185 1.84E-06 -2.003310 0.307493 -1.205621 4 105.3674 18.83258 1.99E-06 -2.135278 0.886541 -1.092146
Series: LRPPI LRPDCP LRPLL LRPCBA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.386339 37.00521 47.21 54.46 At most 1 0.259270 18.44933 29.68 35.65 At most 2 0.117144 7.044829 15.41 20.04 At most 3 0.058986 2.310277 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates no cointegration at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.386339 18.55588 27.07 32.24 At most 1 0.259270 11.40450 20.97 25.52 At most 2 0.117144 4.734552 14.07 18.63 At most 3 0.058986 2.310277 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates no cointegration at both 5% and 1% levels
- 227 -
A 5: Gross Domestic Investments and Banking
VAR Lag Order Selection Criteria Endogenous variables: LRPGDI LRPDCP LRPLL LRPCBA
Lag LogL LR FPE AIC SC HQ
0 -40.58011 NA 0.000150 2.547435 2.725189 2.608795 1 73.32912 195.2730* 5.62E-07* -3.047378* -2.158608* -2.740575* 2 86.34069 19.33147 6.94E-07 -2.876611 -1.276824 -2.324365 3 99.22318 16.19512 9.18E-07 -2.698467 -0.387664 -1.900778 4 115.3974 16.63639 1.12E-06 -2.708425 0.313394 -1.665294
Series: LRPGDI LRPDCP LRPLL LRPCBA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.448070 39.73705 47.21 54.46 At most 1 0.295809 17.15237 29.68 35.65 At most 2 0.080682 3.825543 15.41 20.04 At most 3 0.016413 0.628854 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates no cointegration at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.448070 22.58468 27.07 32.24 At most 1 0.295809 13.32682 20.97 25.52 At most 2 0.080682 3.196690 14.07 18.63 At most 3 0.016413 0.628854 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates no cointegration at both 5% and 1% levels
- 228 -
A 6: Economic Efficiency and Banking
VAR Lag Order Selection Criteria Endogenous variables: REFL LRDCP LRLL LRCBA
Lag LogL LR FPE AIC SC HQ
0 58.01334 NA 4.90E-07 -3.177255 -2.997683 -3.116016 1 149.1783 155.5167 5.94E-09 -7.598722 -6.700863* -7.292526 2 170.6511 31.57765* 4.49E-09* -7.920652 -6.304505 -7.369499* 3 184.7291 17.39045 5.63E-09 -7.807591 -5.473158 -7.011483 4 204.3917 19.66260 5.72E-09 -8.023038* -4.970317 -6.981973 5 217.9918 10.40009 1.02E-08 -7.881869 -4.110861 -6.595847
Series: REFL LRDCP LRLL LRCBA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.835246 97.08453 47.21 54.46 At most 1 0.420632 28.55897 29.68 35.65 At most 2 0.165970 7.817898 15.41 20.04 At most 3 0.023957 0.921446 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.835246 68.52556 27.07 32.24 At most 1 0.420632 20.74107 20.97 25.52 At most 2 0.165970 6.896452 14.07 18.63 At most 3 0.023957 0.921446 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 1 cointegrating equation(s) at both 5% and 1% levels
- 229 -
A 7: Economic Activities and Banking ( seven variables)
VAR Lag Order Selection Criteria Endogenous variables: LRY LRPPI LRPK RPEF LRPDCP LRPLL LRPCBA
Lag LogL LR FPE AIC SC HQ
0 147.6083 NA 9.55E-13 -7.811573 -7.503667 -7.704106 1 308.1587 249.7450 2.05E-15 -14.00882 -11.54556* -13.14907 2 372.4916 75.05504 1.21E-15 -14.86064 -10.24205 -13.24863 3 482.2799 85.39089* 1.14E-16* -18.23777* -11.46383 -15.87348*
Series: LRY LRPPI LRPK RPEF LRPDCP LRPLL LRPCBA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.946967 226.7583 124.24 133.57 At most 1 ** 0.706317 115.1584 94.15 103.18 At most 2 * 0.490810 68.59865 68.52 76.07 At most 3 0.432922 42.95112 47.21 54.46 At most 4 0.314522 21.39528 29.68 35.65 At most 5 0.110426 7.045023 15.41 20.04 At most 6 0.066097 2.598524 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 3 cointegrating equation(s) at the 5% level Trace test indicates 2 cointegrating equation(s) at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.946967 111.5999 45.28 51.57 At most 1 ** 0.706317 46.55972 39.37 45.10 At most 2 0.490810 25.64753 33.46 38.77 At most 3 0.432922 21.55584 27.07 32.24 At most 4 0.314522 14.35025 20.97 25.52 At most 5 0.110426 4.446499 14.07 18.63 At most 6 0.066097 2.598524 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 2 cointegrating equation(s) at both 5% and 1% levels
- 230 -
A 8: Economic Activities versus Banking (six variables)
VAR Lag Order Selection Criteria Endogenous variables: LRY LRPK RPEF LRPDCP LRPLL LRPCBA
Lag LogL LR FPE AIC SC HQ
0 127.6408 NA 4.68E-11 -6.757821 -6.493901 -6.6657061 269.0880 227.8873 1.38E-13 -12.61600 -10.76856* -11.971202 305.2806 46.24608 1.62E-13 -12.62670 -9.195743 -11.429203 361.1080 52.72590* 8.95E-14* -13.72822* -8.713748 -11.97804
Series: LRY LRPK RPEF LRPDCP LRPLL LRPCBA Lags interval (in first differences): 1 to 2
Unrestricted Cointegration Rank Test Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.770540 151.7764 94.15 103.18 At most 1 ** 0.731700 98.78347 68.52 76.07 At most 2 * 0.481354 51.42003 47.21 54.46 At most 3 0.348884 27.78479 29.68 35.65 At most 4 0.265538 12.33839 15.41 20.04 At most 5 0.033540 1.228163 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 3 cointegrating equation(s) at the 5% level Trace test indicates 2 cointegrating equation(s) at the 1% level
- 231 -
A 9: Economic Activities and Banking (six variables, weak exogeneity test)
Tests of cointegration restrictions:
Hypothesized Restricted LR Degrees of No. of CE(s) Log-likehood Statistic Freedom Probability
3 325.5153 43.40068 9 0.000002 3 Cointegrating Equation(s): Convergence achieved after 29 iterations.
Restricted cointegrating coefficients (std.err. in parentheses) LRY LRPK RPEF LRPDCP LRPLL LRPCBA
1.000000 0.434862 0.000000 -0.463129 -0.018567 0.000000 (0.00000) (0.00914) (0.00000) (0.04601) (0.00830) (0.00000) 3.033673 1.000000 0.000000 -1.444269 0.000000 0.059142 (0.02368) (0.00000) (0.00000) (0.13015) (0.00000) (0.02552) 0.000000 0.000000 1.000000 -0.510096 0.065623 0.146774 (0.00000) (0.00000) (0.00000) (0.13601) (0.06847) (0.07474)
Adjustment coefficients (std.err. in parentheses)
D(LRY) -3.917719 1.185396 -0.088618 (3.14400) (1.19344) (0.46307)
D(LRPK) -10.22215 3.511853 -0.141025 (4.16345) (1.58042) (0.61323)
D(RPEF) -4.604478 1.516690 -0.373238 (2.76046) (1.04785) (0.40658)
D(LRPDCP) 0.000000 0.000000 0.000000 (0.00000) (0.00000) (0.00000)
D(LRPLL) 0.000000 0.000000 0.000000 (0.00000) (0.00000) (0.00000)
D(LRPCBA) 0.000000 0.000000 0.000000 (0.00000) (0.00000) (0.00000)
- 232 -
A 10: Economic Activities versus Banking (six variables, Granger causality test)
VEC Pairwise Granger Causality/Block Exogeneity Wald Tests Included observations: 36
Dependent variable: D(LRY)
Exclude Chi-sq df Prob.
D(LRPK) 7.134110 2 0.0282 D(RPEF) 4.150262 2 0.1255
D(LRPDCP) 7.876483 2 0.0195 D(LRPLL) 14.69964 2 0.0006
D(LRPCBA) 16.36040 2 0.0003
All 72.45405 10 0.0000
Dependent variable: D(LRPK)
Exclude Chi-sq df Prob.
D(LRY) 4.540223 2 0.1033 D(RPEF) 4.354045 2 0.1134
D(LRPDCP) 0.479771 2 0.7867 D(LRPLL) 3.249150 2 0.1970
D(LRPCBA) 1.871744 2 0.3922
All 10.92980 10 0.3630
Dependent variable: D(RPEF)
Exclude Chi-sq df Prob.
D(LRY) 6.971076 2 0.0306 D(LRPK) 10.03194 2 0.0066
D(LRPDCP) 11.15220 2 0.0038 D(LRPLL) 17.10188 2 0.0002
D(LRPCBA) 20.85605 2 0.0000
All 91.24905 10 0.0000
Dependent variable: D(LRPDCP)
Exclude Chi-sq df Prob.
D(LRY) 2.545641 2 0.2800 D(LRPK) 2.550384 2 0.2794 D(RPEF) 1.857810 2 0.3950 D(LRPLL) 1.652477 2 0.4377
D(LRPCBA) 1.243866 2 0.5369
All 8.092034 10 0.6198
Dependent variable: D(LRPLL)
Exclude Chi-sq df Prob.
D(LRY) 0.508142 2 0.7756 D(LRPK) 0.198049 2 0.9057 D(RPEF) 0.609144 2 0.7374
D(LRPDCP) 3.428993 2 0.1801 D(LRPCBA) 2.901985 2 0.2343
All 8.299474 10 0.5996
Dependent variable: D(LRPCBA)
Exclude Chi-sq df Prob.
D(LRY) 2.071361 2 0.3550 D(LRPK) 0.278634 2 0.8700 D(RPEF) 2.150186 2 0.3413
D(LRPDCP) 6.615526 2 0.0366 D(LRPLL) 8.918884 2 0.0116
All 16.79589 10 0.0790
- 233 -
A 11: Co-integration Rank Test Among Economic Variables (4 variables)
VAR Lag Order Selection Criteria Endogenous variables: LRY LRPPI LRPK RPEF
Lag LogL LR FPE AIC SC HQ
0 135.6587 NA 5.09E-09 -7.744631 -7.565059 -7.683391 1 204.3688 117.2113 2.31E-10 -10.84522 -9.947365* -10.53903 2 220.6325 23.91718 2.38E-10 -10.86073 -9.244588 -10.30958 3 243.1009 27.75510* 1.82E-10 -11.24123 -8.906796 -10.44512 4 261.5533 18.45238 1.98E-10 -11.38549 -8.332766 -10.34442 5 286.4934 19.07181 1.81E-10* -11.91137* -8.140365 -10.62535*
Series: LRY LRPPI LRPK RPEF Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.857906 132.4700 47.21 54.46 At most 1 ** 0.633710 58.32194 29.68 35.65 At most 2 ** 0.348376 20.15737 15.41 20.04 At most 3 * 0.097123 3.882423 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 4 cointegrating equation(s) at the 5% level Trace test indicates 3 cointegrating equation(s) at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.857906 74.14807 27.07 32.24 At most 1 ** 0.633710 38.16456 20.97 25.52 At most 2 * 0.348376 16.27495 14.07 18.63 At most 3 * 0.097123 3.882423 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 4 cointegrating equation(s) at the 5% level Max-eigenvalue test indicates 2 cointegrating equation(s) at the 1% level
- 234 -
A 11(continued): Co-integration Rank Test Among Economic Variables
Series: LRY LRPPI LRPK RPEF Lags interval (in first differences): 1 to 4
Unrestricted Cointegration Rank Test Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.688538 84.10553 47.21 54.46 At most 1 ** 0.567962 44.44531 29.68 35.65 At most 2 * 0.287589 15.91111 15.41 20.04 At most 3 * 0.120914 4.381681 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 4 cointegrating equation(s) at the 5% level Trace test indicates 2 cointegrating equation(s) at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.688538 39.66022 27.07 32.24 At most 1 ** 0.567962 28.53420 20.97 25.52 At most 2 0.287589 11.52943 14.07 18.63
At most 3 * 0.120914 4.381681 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 2 cointegrating equation(s) at both 5% and 1% levels
- 235 -
A 12: Co-integration Rank Test Among Economic Variables (3 variables)
VAR Lag Order Selection Criteria Endogenous variables: LRY LRPK RPEF
Lag LogL LR FPE AIC SC HQ
0 118.9879 NA 2.19E-07 -6.822816 -6.688138 -6.776887 1 174.9125 98.69056 1.39E-08 -9.583090 -9.044375* -9.399373 2 188.4827 21.55254* 1.08E-08* -9.851921* -8.909169 -9.530415* 3 192.1032 5.111318 1.53E-08 -9.535480 -8.188692 -9.076187 4 203.7356 14.36949 1.40E-08 -9.690330 -7.939505 -9.093249 5 213.5544 10.39632 1.51E-08 -9.738492 -7.583630 -9.003622
Series: LRY LRPK RPEF Lags interval (in first differences): 1 to 1
Unrestricted Cointegration Rank Test Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.309359 23.00830 29.68 35.65 At most 1 0.152819 9.313282 15.41 20.04 At most 2 0.082286 3.177165 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates no cointegration at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.309359 13.69501 20.97 25.52 At most 1 0.152819 6.136117 14.07 18.63 At most 2 0.082286 3.177165 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates no cointegration at both 5% and 1% levels
- 236 -
A 13: Co-integration Rank Test Among the Banking Variables
VAR Lag Order Selection Criteria Endogenous variables: LRPLL LRPDCP LRPCBA
Lag LogL LR FPE AIC SC HQ
0 -19.91878 NA 0.000773 1.348164 1.482843 1.394093 1 53.28014 129.1746 1.78E-05* -2.428244* -1.889528* -2.244526* 2 58.96423 9.027669 2.19E-05 -2.233190 -1.290438 -1.911685 3 63.36558 6.213673 2.98E-05 -1.962681 -0.615893 -1.503388 4 77.42185 17.36363* 2.37E-05 -2.260109 -0.509284 -1.663027 5 82.81396 5.709294 3.30E-05 -2.047880 0.106982 -1.313011
Series: LRPLL LRPDCP LRPCBA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.290831 18.97298 29.68 35.65 At most 1 0.093708 5.913859 15.41 20.04 At most 2 0.055628 2.174916 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates no cointegration at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.290831 13.05912 20.97 25.52 At most 1 0.093708 3.738943 14.07 18.63 At most 2 0.055628 2.174916 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates no cointegration at both 5% and 1% levels
- 237 -
A 14: Broaddus’ Model
VAR Lag Order Selection Criteria Endogenous variables: LR DR
Lag LogL LR FPE AIC SC HQ
0 -160.5661 NA 48.76398 9.562712 9.652498 9.593332 1 -116.7499 79.90021 4.691472* 7.220580* 7.489938* 7.312439* 2 -114.0375 4.626984 5.077735 7.296323 7.745253 7.449421 3 -111.8609 3.456990 5.696019 7.403581 8.032082 7.617918 4 -110.7578 1.622100 6.848719 7.573991 8.382064 7.849567 5 -103.1043 10.35486* 5.649338 7.359074 8.346719 7.695889
* indicates lag order selected by the criterion Series: LR DR Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.332763 16.90408 15.41 20.04 At most 1 0.039436 1.528903 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at the 5% level Trace test indicates no cointegration at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.332763 15.37518 14.07 18.63 At most 1 0.039436 1.528903 3.76 6.65
- 238 -
A15: Demand for and Supply of Bank Credit (five variables)
Endogenous variables: LDCP LR DR LLC LGDP VAR Lag Order Selection Criteria
Lag LogL LR FPE AIC SC HQ
0 -188.2637 NA 0.043038 11.04364 11.26583 11.12034 1 -28.44384 264.8443 1.97E-05 3.339648 4.672803* 3.799853 2 4.035302 44.54282* 1.41E-05* 2.912268 5.356387 3.755978 3 29.12982 27.24547 1.81E-05 2.906868 6.461949 4.134081 4 67.48059 30.68062 1.49E-05 2.143967* 6.810011 3.754684*
Series: LDCP LR DR LLC LGDP Lags interval (in first differences): 1 to 3
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.844030 143.6109 68.52 76.07 At most 1 ** 0.791747 78.57776 47.21 54.46 At most 2 0.350623 23.66275 29.68 35.65 At most 3 0.210831 8.551785 15.41 20.04 At most 4 0.007533 0.264663 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 2 cointegrating equation(s) at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.844030 65.03310 33.46 38.77 At most 1 ** 0.791747 54.91501 27.07 32.24 At most 2 0.350623 15.11096 20.97 25.52 At most 3 0.210831 8.287123 14.07 18.63 At most 4 0.007533 0.264663 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 2 cointegrating equation(s) at both 5% and 1% levels
- 239 -
A 16: Demand for and Supply of Bank Credit (six variables)
VAR Lag Order Selection Criteria Endogenous variables: LDCP LR DR LLC LGDS LGDP
Lag LogL LR FPE AIC SC HQ
0 -309.8633 NA 2.779193 18.04933 18.31596 18.14137 1 -138.0913 274.8351 0.001227 10.29093 12.15735* 10.93522 2 -96.85851 51.83553* 0.001101 9.991915 13.45812 11.18845 3 -45.30272 47.13673 0.000793* 9.103012* 14.16900 10.85179*
Series: LDCP LR DR LLC LGDS LGDP Lags interval (in first differences): 1 to 2
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.776818 157.2689 94.15 103.18 At most 1 ** 0.696054 104.7771 68.52 76.07 At most 2 ** 0.611682 63.09541 47.21 54.46 At most 3 * 0.425108 29.98780 29.68 35.65 At most 4 0.259140 10.61274 15.41 20.04 At most 5 0.003273 0.114737 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 4 cointegrating equation(s) at the 5% level Trace test indicates 3 cointegrating equation(s) at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.776818 52.49184 39.37 45.10 At most 1 ** 0.696054 41.68168 33.46 38.77 At most 2 ** 0.611682 33.10761 27.07 32.24 At most 3 0.425108 19.37506 20.97 25.52 At most 4 0.259140 10.49800 14.07 18.63 At most 5 0.003273 0.114737 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 3 cointegrating equation(s) at both 5% and 1% levels
- 240 -
A 17: Money demand Function
VAR Lag Order Selection Criteria Endogenous variables: LRM LRGDP LI ID
Lag LogL LR FPE AIC SC HQ
0 -149.7412 NA 0.099476 9.043601 9.223173 9.104841 1 -50.63730 169.0596* 0.000756* 4.155135 5.052995* 4.461331* 2 -37.23285 19.71243 0.000919 4.307815 5.923961 4.858967 3 -27.11021 12.50444 0.001453 4.653542 6.987976 5.449651 4 -2.022319 25.08789 0.001072 4.118960 7.171681 5.160025 5 25.17061 20.79459 0.000857 3.460553* 7.231561 4.746574
Series: LRM LRGDP LI ID Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.482521 49.60758 47.21 54.46 At most 1 0.349654 24.57367 29.68 35.65 At most 2 0.187976 8.224160 15.41 20.04 At most 3 0.008166 0.311579 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at the 5% level Trace test indicates no cointegration at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None 0.482521 25.03392 27.07 32.24 At most 1 0.349654 16.34951 20.97 25.52 At most 2 0.187976 7.912582 14.07 18.63 At most 3 0.008166 0.311579 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates no cointegration at both 5% and 1% levels
- 241 -
A 18: Transaction Costs Function
VAR Lag Order Selection Criteria Lag LogL LR FPE AIC SC HQ
0 -106.6102 NA 0.126701 6.447658 6.582337 6.493587 1 41.74884 261.8100* 3.50E-05 -1.749932 -1.211216* -1.566214* 2 48.48847 10.70411 4.06E-05 -1.616969 -0.674216 -1.295463 3 58.49233 14.12310 3.96E-05 -1.676019 -0.329231 -1.216726 4 71.83597 16.48332 3.29E-05* -1.931528 -0.180702 -1.334446 5 81.79145 10.54110 3.50E-05 -1.987733* 0.167129 -1.252863
Series: ID LRGDP LCPI Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.765924 59.19774 29.68 35.65 At most 1 0.088514 4.017598 15.41 20.04 At most 2 0.012962 0.495787 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at both 5% and 1% levels
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.765924 55.18014 20.97 25.52 At most 1 0.088514 3.521810 14.07 18.63 At most 2 0.012962 0.495787 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 1 cointegrating equation(s) at both 5% and 1% levels
- 242 -
A 19: Schumpeter’s Credit
VAR Lag Order Selection Criteria Endogenous variables: LRPI LRDCP LRLC LI
Lag LogL LR FPE AIC SC HQ
0 -112.7726 NA 0.011306 6.868975 7.048547 6.930214 1 -30.97320 139.5401* 0.000238* 2.998424 3.896283* 3.304619* 2 -21.48735 13.94978 0.000364 3.381609 4.997755 3.932761 3 -9.053977 15.35888 0.000502 3.591410 5.925844 4.387519 4 6.507175 15.56115 0.000649 3.617225 6.669946 4.658290 5 33.63637 20.74586 0.000521 2.962566* 6.733575 4.248588
* indicates lag order selected by the criterion Series: LRPI LRDCP LRLC LI Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.523685 50.12725 47.21 54.46 At most 1 0.344404 21.94353 29.68 35.65 At most 2 0.115774 5.899499 15.41 20.04 At most 3 0.031694 1.223877 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 1 cointegrating equation(s) at the 5% level Trace test indicates no cointegration at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None * 0.523685 28.18372 27.07 32.24 At most 1 0.344404 16.04403 20.97 25.52 At most 2 0.115774 4.675622 14.07 18.63 At most 3 0.031694 1.223877 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 1 cointegrating equation(s) at the 5% level Max-eigenvalue test indicates no cointegration at the 1% level
- 243 -
A 20: Monetary Competence
VAR Lag Order Selection Criteria Endogenous variables: CDC CNFA
Lag LogL LR FPE AIC SC HQ
0 -288.4149 NA 89995.10 17.08323 17.17302 17.11385 1 -216.3807 131.3564* 1646.466* 13.08122* 13.35058* 13.17308* 2 -214.5260 3.163894 1874.236 13.20741 13.65634 13.36051 3 -213.2798 1.979313 2220.722 13.36940 13.99790 13.58374 4 -212.4467 1.225116 2712.866 13.55569 14.36376 13.83127 5 -210.4599 2.688112 3123.079 13.67411 14.66175 14.01093
Series: CDC CNFA Lags interval (in first differences): No lags
Unrestricted Cointegration Rank Test
Hypothesized Trace 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.498585 32.64607 15.41 20.04 At most 1 * 0.155311 6.413903 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Trace test indicates 2 cointegrating equation(s) at the 5% level Trace test indicates 1 cointegrating equation(s) at the 1% level
Hypothesized Max-Eigen 5 Percent 1 Percent No. of CE(s) Eigenvalue Statistic Critical Value Critical Value
None ** 0.498585 26.23217 14.07 18.63 At most 1 * 0.155311 6.413903 3.76 6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level Max-eigenvalue test indicates 2 cointegrating equation(s) at the 5% level Max-eigenvalue test indicates 1 cointegrating equation(s) at the 1% level
- 244 -
Appendix B: Research Data
year POP Labor GDP GDI GI PI DC NFA
1964 0,4 0,227 56,17 10,60 4,12 6,48 5,46 6 1965 0,41 0,229 62,69 11,83 4,60 7,23 5,99 6,79 1966 0,42 0,232 69,70 13,15 5,12 8,04 4,49 11,12 1967 0,43 0,235 76,68 14,47 5,63 8,84 4,46 10,5 1968 0,44 0,238 61,94 11,69 4,55 7,14 7,13 8,96 1969 0,45 0,242 77,36 14,60 5,68 8,92 7,97 10 1970 0,47 0,245 81,85 15,45 6,01 9,44 8,19 16,31 1971 0,48 0,248 81,42 15,36 5,98 9,39 10,83 15,39 1972 0,5 0,248 108,79 20,53 7,99 12,54 12,06 23,68 1973 0,52 0,248 109,77 20,71 8,06 12,66 30,35 28,93 1974 0,54 0,248 158,35 29,88 11,62 18,26 38,26 51,75 1975 0,56 0,248 221,20 41,74 16,24 25,50 45,11 61,07 1976 0,57 0,251 278,33 52,52 20,43 32,09 72,48 49,91 1977 0,59 0,255 355,12 67,01 26,07 40,95 82,69 39,93 1978 0,61 0,258 360,66 68,06 26,47 41,58 117,41 21,72 1979 0,63 0,262 425,00 80,20 31,20 49,00 158,68 -22,71 1980 0,65 0,278 435,00 82,08 31,93 50,16 195,59 -44,96 1981 0,67 0,282 476,00 89,82 34,94 54,88 235,52 -77,72 1982 0,7 0,286 522,00 98,50 38,31 60,19 291,41 -117,69 1983 0,72 0,291 606,00 135,14 56,36 78,78 370,03 -177,79 1984 0,75 0,295 618,00 140,29 59,95 80,34 428,29 -311,77 1985 0,77 0,309 782,00 165,78 93,06 72,73 527,56 -338,05 1986 0,8 0,314 1085,00 167,09 77,04 90,06 363,21 -449,44 1987 0,83 0,319 1486,00 291,26 102,53 188,72 205,7 -338,74 1988 0,87 0,324 1636,00 253,58 109,61 143,97 241,64 -221,13 1989 0,9 0,330 1942,00 341,79 102,93 238,87 216,32 -204,8 1990 0,94 0,342 2367,00 487,60 120,72 366,89 85,14 -13,07 1991 0,97 0,348 2630,00 507,59 134,13 373,46 -0,12 95,4 1992 1,01 0,354 2948,00 560,12 156,24 403,88 278,39 518,83 1993 1,04 0,360 3229,00 629,66 232,49 397,17 230,15 601,79 1994 1,08 0,366 3461,00 553,76 211,12 342,64 199,34 599,37 1995 1,11 0,373 3492,00 817,13 450,47 366,66 281,13 776,05 1996 1,15 0,379 3880,00 787,64 531,56 256,08 305,92 774,28 1997 1,19 0,386 4180,00 618,64 271,70 346,94 413,56 833,7 1998 1,23 0,393 4479,00 824,14 264,26 559,88 484,99 955,55 1999 1,27 0,400 4922,00 876,12 260,87 615,25 637,34 1059,95 2000 1,31 0,426 5382,00 931,09 247,57 683,51 741,32 1351,87 2001 1,35 0,434 6556,00 1166,97 314,69 852,28 1530,96 1194,18 2002 1,39 0,441 7364,00 552,30 1008,87 3127,08 1894,21 1319,55
- 245 -
year GD M LL CBA DCP LC LR DR
1964 0 10,2 3,3 10,26 9,93 10,27 7,36 4,61 1965 1,76 10,84 2,76 12,49 12,23 12,49 9,44 5,91 1966 0,76 12,76 5,11 13,58 12,76 13,58 9,74 6,10 1967 0,23 12,7 6,77 15,19 14,26 15,19 9,30 5,82 1968 0,73 13,66 11,42 17,62 15,28 17,62 11,32 7,09 1969 0,3 14,08 13,41 19,49 16,21 19,49 12,20 7,64 1970 0,03 16,46 13,61 20,87 15,22 20,87 11,21 7,02 1971 0,01 16,38 9,53 17,77 15,37 17,77 8,91 5,58 1972 0,5 22,75 14,81 20,08 12,5 20,08 8,79 5,51 1973 0 31,84 30,19 37,87 20,94 37,87 14,84 9,29 1974 1,43 33,04 32,2 47,6 24,08 47,6 18,16 11,37 1975 0,75 37,35 40,1 56,41 22,96 56,41 16,25 10,18 1976 7,14 49,43 49,9 83,53 39,12 83,53 17,82 11,16 1977 25,99 39,8 37,87 85,76 53,74 85,76 12,23 7,66 1978 29,14 57,78 56,14 113,98 75,4 113,98 15,00 10,00 1979 46,67 57,78 67,11 132,53 83,06 132,53 15,00 10,00 1980 29,56 61,28 110,2 212,99 98,46 212,99 15,00 10,00 1981 63,76 76,97 134,87 233,49 104,44 233,49 18,00 9,50 1982 51,22 87,22 161,21 274,32 102,91 274,32 18,00 9,50 1983 106,35 100,32 173,16 295,77 137,59 295,77 18,00 9,50 1984 131,15 99,58 174,14 276,33 158,17 276,33 18,00 9,00 1985 179,41 162,1 228,8 334,67 212,37 334,67 14,48 4,73 1986 81,2 166,49 218,13 386,78 192,34 386,78 28,00 11,87 1987 93,39 197,83 189,67 360,72 184,14 360,72 27,92 12,17 1988 118,22 213,52 186,71 404,45 208,53 404,45 29,54 14,54 1989 128,28 260,25 181,86 452,03 231,31 452,03 26,83 13,91 1990 164,28 296,19 195,07 483,41 265,45 483,41 26,50 15,17 1991 154,4 393,58 268,53 582,81 300,52 582,81 26,50 13,79 1992 427,28 435,17 232,05 590,99 222,85 590,99 26,75 12,92 1993 406,46 461,38 240,85 626,26 361,9 626,26 26,08 13,08 1994 398,43 407,53 204,13 617,54 385,44 617,54 25,00 12,42 1995 481,65 471,47 227,46 742,6 342,35 742,6 25,04 12,54 1996 615,43 453,49 202,42 860,12 341,89 860,12 25,50 13,00 1997 688,09 629,41 272,86 1116 425,26 1116 25,50 13,00 1998 756,52 626,57 282,98 1260,19 489,76 1260,19 25,38 12,88 1999 827,25 716,04 340,28 1536,46 591,41 1536,46 24,00 11,50 2000 1042,93 983,53 447,23 1783,67 652,18 1783,67 24,00 11,50 2001 1329,51 1125,41 528,62 2293,15 873,21 2293,15 24,00 11,50 2002 1756,77 963,36 1203,05 1866,29 3127,08 3127,08 24,00 11,29
- 246 -
year CA EX OIP CPI
1964 -5,21 533,27 186,64 4,99 1965 -4,57 533,98 206,54 5,06 1966 -4,10 533,4 186,64 5,07 1967 -3,05 524,51 179,28 5,14 1968 -17,04 457,11 166,06 5,35 1969 -6,31 456,47 206,77 5,62 1970 -2,31 457,54 228,18 5,51 1971 -16,09 466,74 250,91 5,68 1972 -12,15 477,76 253,95 6,17 1973 -13,41 566,32 391,31 6,60 1974 -7,33 558,34 739,06 7,21 1975 -31,12 530,35 432,96 9,08 1976 -89,65 431,15 422,99 10,62 1977 -67,52 416,66 546,86 11,94 1978 -56,55 458,19 630,93 13,00 1979 -58,25 506,42 562,74 13,80 1980 -202,28 555,29 1236,58 14,67 1981 -60,45 484,07 1257,42 15,83 1982 -14,09 417,86 827,5 17,14 1983 -9,09 362,11 965,33 18,73 1984 -39,55 272,3 836,58 21,65 1985 -25,81 247,55 675,75 26,37 1986 -7,60 139,82 993,58 35,60 1987 -31,21 135,14 758,5 52,05 1988 63,80 142,62 935,75 58,49 1989 36,90 126,57 817,5 64,81 1990 -63,91 121,49 1325,5 71,42 1991 -2,63 109,37 1237,83 77,92 1992 -2,95 107,15 799,13 87,28 1993 -148,53 104,67 1092,15 92,47 1994 -93,45 99,66 954,8 96,13 1995 -150,16 100 909,92 100,00 1996 -508,28 97,55 962 104,75 1997 -154,66 93,64 988,42 107,99 1998 -107,50 89,72 988,75 109,19 1999 -137,82 83,83 834,74 113,35 2000 -166,84 75,02 843,93 114,31 2001 -26,22 60,96 833,16 119,44 2002 -405,02 48,42 753,29 129,73
The data are in millions of Gambian Dalasis, except of course the LR, DR and CPI; and OIP is an oil price index, and EX is the exchange rate index. The definitions of the variables used in the research are found in section 4.5, Chapter 4.
- 247 -
Appendix C : Survey Questionnaire
I am a PhD student from the University of Kiel, Germany. This questionnaire is part of my dissertation that looks at the relationships between banking and economic growth in the Gambia under the supervision of Professor Dr. H.-W. Wohltmann and Professor Dr. Helmut Herwartz. I would appreciate if you could take some of your precious time and fill in this questionnaire.
1. Sex [ ] Male [ ] female 2. Education ……………… 3. Age [ ] 18 - 28 years [ ] 29 - 39 years [ ] 40 – 50 years [ ] above 50 years. 4. Please, tick the bank accounts you hold,
i. Savings Account ii. Time Deposits iii. Current Account iv. Others
5. Why do you place your money in the banks? ( Please, tick on the order of importance to you from 1 to 5, with 1 indicating the most important and 5 the least important)
[ ] To avoid the risk of theft, loss and destruction [ ] To earn interest income [ ] To qualify for loan facility [ ] To make easy payments [ ] To save for the future 6. How do you rate your chance of obtaining a bank loan for the following purposes
(give percentages from 0% to 100%, with 0% meaning no chance at all and 100% meaning certain).
i. Business/ Project ……… ………………% ii. House purchase / construction ………….% iii. Land purchase……………………………% iv. Consumption……………………………..% v. Agriculture……………………………….% vi. Education………………………………...%
7. I have obtained a bank loan because (please, tick the one most relevant to you)
[ ] I have a good project / business [ ] I have good collateral [ ] I have a good income stream [ ] I have a good connection in the bank [ ] I have ……………………………..
8. rate the services of your bank on the scale of 1 to 7, with 1 indicating very poor and 7 excellent. [ ] 1 [ ] 2 [ ] 3 [ ] 4 [ ] 5 [ ] 6 [ ] 7
- 248 -
Eidesstattliche Erklaerung
Ich versichere hiermit an Eides Statt, dass ich die Dissertation
“Banking and Economic Growth: Case of the Gambia“ selbstaendig angefertigt habe, alle von anderen Autoren woertlich uebernommenen Stellen
und sich an die gedankengaenge anderer Autoren eng anlehnende Ausfuehrungen
besonders gekennzeichnet und keine anderen als die in der Dissertationsschrift
angegebenen Hilfsmittel benutzt habe.
Kiel, den 20. Juli 2005
Bukhari M. S. Sillah
- 249 -
Curriculum Vitae Bukhari M. S. Sillah Wassu, Nyani District, C.R.D The Gambia [email protected], Birth date : 9th August 1973.
Birth place: Wassu, Nyani, C.R.D. The Gambia.
Academic Qualifications:
1999 – 2001 : Master of Economics, International Islamic University, Malaysia. 1994 – 1998 : Bachelor of Business Administration, International Islamic University, Malaysia. 1998 : Certificates of Malaysian Institute of Futures and Options 1990 – 1993 : Technical Secondary School Certificate, African Islamic Center, Khartoum, Sudan.
Awards: • African Muslim Agency Award and the International Islamic University Award
for the study of Master of Economics. • Malaysian Technical Cooperation Programme Award and the International
Islamic University Award for the study of Bachelor of Business Administration. • Second Best Student Award at the Department of Business Administration,
International Islamic University, Malaysia. Work Experiences: 2001 – 2002 : Economics Lecturer, University of the Gambia. 1999 – 2001 : Graduate Teaching Assistant, International Islamic University, Malaysia. 1998 – 1999 : Cadet Administration Officer, Personnel Management Office, Government of the Gambia. Languages: English, Arabic, some Deutsch, Mandingo, Fulani and some Wolof. Computer Languages: Word, Excel, PowerPoint, Shazam, E-Views and Scientific Work. Signature:………..………………………………….
